tendre.AI
Free Trial

Changelog

Every update to tendre.AI, the local-first AI image and video app, version by version.

v0.9.0

  • Create an image: Generate an image and Multi-Ref Image are now one screen, Create an image. Text alone is enough, references are added on top. Free LoRAs (outside character LoRAs) and batch queue disappear: number of images replaces them. Old Gallery images reopen there with their settings. Create a video is the startup screen.
  • Create an image and Edit: one default image, and your choice of 1, 2 or 4 images (8 is no longer offered), like everywhere you generate an image.
  • Edit: drag an image anywhere on screen, from file explorer or app, to load it. If a photo is already there, the screen switches on its own to Two images and the new one becomes image 2.
  • Elements: for a movement, you choose Choreography or Camera first, then give only a name and the video.
  • Retoucher is now called Edit (Inpainting in English), in the menu and at the top of the screen. The screen does not change: a prompt, a painted area if you want, one or two images.
  • Character sheet: Import a character sheet takes a sheet made elsewhere. A single image (your full board) is kept as is and serves as reference in Multi-Ref image and video; multiple images (up to 7) go into the boxes, in order of their file names. You can also drag your images from file explorer onto the character sheets section, or just one onto an empty box.
  • Character sheet: the Shoes view is now called Feet, and its text no longer talks about hair (it was falling to the feet in the frame).
  • Removal: the Library > LoRA screen (the table of all LoRA files) disappears. Your LoRAs and characters stay put; Recreate the character from an orphaned LoRA no longer exists.
  • Removal: the Generation size area in Settings disappears; quality (Small, Standard, High 2K) is chosen directly when you create an image.
  • Removal: the Upscale screen (doubling an already-made image resolution) disappears. For a larger image, choose High quality (2K) when you create it.
  • Create a video: Ultra starts at 10 s in 480p when you pick it. You then adjust duration and resolution as before.
  • Create a video: if you change duration then go back through the AI button, the prompt is rewritten for the new duration, from your original idea. Before, it kept the old duration breakdown.
  • Render queue: all renders go through it, on your card and online, the same way: chat, Convert to HD, Face swap, Add sound, sending a video LoRA and film export, which does not lock the card. Each line says what is rendering, with the prompt or character next to it, its time and progress. A job in multiple stages (starting image then clip) makes one line, and Cancel only appears where it actually stops the render.
  • Create a video: an online render appears in the queue at bottom left, with its progress, like other renders.
  • Fix: canceling a video LoRA training stops the training on our servers. On a PC able to train locally, Cancel could stop a local training that did not exist, and online training kept going.
  • Lucas is again described as a man in Generate, Multi-Refs, Create a video and his character sheet. Since 0.8.7, his age added to his sheet made him described as a woman.
  • Characters: the views of a character sheet always start from the full-length front view (body) or bust (face). A dropped photo is no longer redressed, and Redo gives a new image with each click.
  • Characters: the profile view of a character sheet no longer sometimes shows the character twice side by side.
  • Characters: you can launch several character sheet views in a row, on one or more characters. They go one after another, and the render queue shows them as pending.
  • Multi-Ref online: a very elongated reference image (a panorama, a car in profile) is no longer rejected. It is completed by bands of its color, without being cropped.
  • Fix: Create a video sends all images of an object, animal or location, gathered into one image. Before, an object went only with its first three photos, and a location with its first, with nothing to say so.
  • Elements: an object, location or animal can have up to 6 images instead of 5.
  • Create a video: a Movement (a 4 to 30 s video) gives the render, at your choice, its camera movement (framing and rhythm included) or its choreography (body gestures, same rhythm), without taking its content. The choice is made on the Movement, in Elements. One movement per video, online with the Ultra model; its price is added to the second, same for both.
  • Fix: importing a character checks the file names it contains and always sorts them into the app folders.
  • Library, Elements: an object, location or animal is shared like a character. Share on its card saves it in a file (.tendreobject, .tendrelocation or .tendreanimal) with its images and description; Import adds it to the Library of the person who receives it.
  • Characters: the Import button also opens a shared object, location or animal, and sorts it into the Library. Old .tendrechar files import as before.
  • Starter mode leaves the computer: everyone moves to Studio, on Windows and Mac alike. A Starter profile opens in Studio on Create a video, in Cloud GPU, with a banner saying so. The setup assistant no longer asks for mode; without a capable card, it no longer asks for roleplay either (on by default, you turn it off in Settings). On phone, Starter stays as the app screen.
  • Create a video becomes a project (Windows and Mac): a sequence of shots, each rendered separately, sorted in its project. An Untitled Project opens right away and gets a few-word title on its first render. The project manager creates, duplicates and deletes (with Undo). The first rendered shot sets the project mode, format, resolution, frame rate and sound. On phone, the screen stays as before.
  • Create a video: a timeline at the bottom of the screen (key T), with FREE tracks: any video track takes shots, Gallery clips, images and titles; any audio track takes sound from a shot, music and sounds. As many tracks as you want, to rename, hide or remove. Shots are numbered in film order; the bottom video track is the base (empty = black and silence), those above are overlaid. A shot's sound follows its image, with its J and L cuts, volume and fades. Between two images: hard cut or fade to black; between two joined sounds: crossfade. You cut (S), drag (an element that crosses the middle of another swaps places with it), link clips in a group, and everything undoes (Ctrl+Z, Ctrl+Shift+Z).
  • Create a video: a floating monitor plays the timeline like export, fades and sounds included. You move it, resize it, fullscreen it or detach it to a second screen; its controls: play, frame by frame, loop, volume.
  • Create a video: Export assembles the film in the folder and under the name of your choice (the same name replaces the old one), then sorts it into the Gallery. The Edit screen disappears: everything happens in the timeline.
  • Create a video: export offers MP4 or MOV. MOV keeps source quality (4:4:4 color, uncompressed sound), MP4 is high quality; same size, same frame rate, same sound.
  • Create a video: drag a video, image or sound from your file explorer onto a timeline track: it lands at the drop spot, a video as a clip. Browse, in the Add window, also opens your files.
  • Create a video: a video, image or sound added from your files is LINKED, never copied: it stays in its place on your disk. If it moved or disappeared, the app says so (not found) and Link lets you find it; export waits for it to be linked. In Sounds, an imported sound is marked, and deleting a sound used by a timeline asks for confirmation.
  • Create a video: in the timeline, scroll wheel moves through time, Shift + scroll zooms under the pointer and Alt + scroll changes track height.
  • Create a video: from the second shot on, settings set by the first rendered shot fit on one line with a lock, and a link opens that reference shot.
  • Create a video: in Add a reference and around the shot, a movement shows an image from its video and its mode (Choreography or Camera). A video outside the engine limits is rejected before sending, with the limit and your value; nothing is charged.
  • Create a video: an (i) next to references reminds you that any lit reference goes to render, even without @, and you turn off a bubble to remove it.
  • Create a video: the generator card expands by dragging its left, right and bottom edges; the size you chose is kept for all your shots and all your projects.
  • Create a video: the prompt assistant keeps your lines word for word and invents none; each shot and the sound part are on their own paragraph.
  • Create a video: tracking an online render does not drop after a network cut; it retries until it gets your clip back.
  • First launch: when your graphics card does not allow creating on your computer, the screen says so simply: everything is done online.
  • Sounds: Import a folder adds all sounds from a folder at once, and Add a sound keeps the file in place: nothing is copied. Sounds, even long music, play right away.
  • Generate an image of a subject or key frame makes one by default (1 credit); 2 or 4 are chosen next to the price.
  • Create a video: in Ultra, up to 30 references per shot (8 characters recommended, 12 at most), around the shot, with Add a reference. In Quality: 4 subjects and one location. An (i) on Quality and Ultra says what each mode does.
  • Create a video: an Ultra render comes out as MP4, as the engine delivers it, with original signature included. Optional, under the sound setting: MOV (4:4:4 image and uncompressed sound), kept for renders that follow.

v0.8.8

  • Face swap: Import an image option in the character selection window. A simple photo serves as a character without creating a card.
  • License: Sign in with my account button at first launch and in the purchase screen. The website opens with a code, you sign in, you choose your license if you have multiple, and the app activates automatically without importing a file. On Windows, Mac and Android.
  • On a computer whose graphics card can't compute locally, Upscale and the LoRA library no longer appear in the menu instead of staying grayed out.
  • Face swap: the first image of a character could show the original person in the video, an image from early in their session. It now shows the character properly.
  • Online video: canceling a render during upload no longer lets a phantom render go through. It stops as soon as the server receives it and nothing is charged.
  • Online image: similarly, canceling an image during upload no longer lets a phantom render go through. It stops as soon as the server receives it and nothing is charged.
  • Voice: an audio file imported into a card or the voice library is at most 20 MB. Only the first 20 seconds are used, a short clip is enough.
  • Characters: with a license, the voice on your card is kept online, encrypted, along with the character's views (their first 20 seconds), so the API and MCP can find it. It leaves when you remove it from the card or when the character is used in a render; if changed or removed, it updates online too, and a shared character takes precedence on import.
  • Characters: the views of a character sheet no longer disappear if you move or empty your images folder. The app keeps its own copy of each view; the one in your folder is now just a copy for your use.
  • Menu: a single Library section at the top for everything you save and reuse, your characters, assets, voices and LoRAs. Assets brings together objects, locations and animals on one page, which had two entries. When creating an asset, Object comes first and is selected by default. In English, the section is called Assets.
  • Chat: when you're away from the character (on the road, on the phone), they no longer believe you've arrived just because you're commenting on a photo or they're inviting you over. The exchange needs to show your arrival, you come in, they open the door.
  • Chat: when the photo shows you standing next to them or sitting further away, you're the one drawn. Before, the character could appear twice in the image.
  • Video and Face swap on your graphics card: when the engine can't read a model anymore due to Windows memory limits, it restarts automatically. Resume starts on a fresh engine without closing the app. In Settings, Clear engine memory also restarts the video engine.
  • Pro plan: 4K video is no longer offered, not in the app, not via API, not via MCP. Video goes up to 1080p everywhere; the license screen and API and connector tabs no longer promise it.
  • Multi-Ref Image: the Add a subject window is the Create a video one. It opens on Characters, then Objects, Animals and Gallery last, with search on the same line.
  • Gallery: no more Root button under Open. In a subfolder, a back arrow goes up one folder.
  • Characters: training photos online go straight to storage instead of passing through our server. Uploading a batch of 40 photos no longer strains the service.
  • Character card: when you scroll the card, the left column (photo, name and steps) stays below the top banner instead of going behind it.
  • Multi-Ref Image and Video, Generate, Face swap: highlighted @mentions stay on their word when scene text exceeds field height or you resize it. Before, highlights would shift to other words.
  • Video: Multi-Ref Video becomes Create a video with a New thumbnail in the menu. The old screen of the same name disappears; its clips come back in Create a video with Resume.
  • Create a video and Starter: Ultra moves to a new engine at per-second pricing (480p, 720p or 1080p, 4 to 30 s, sound included), 24 fps. Your characters, objects and locations each count fully; LoRAs don't apply. A New thumbnail signals it on the Ultra button. The progress bar follows the engine's actual duration.
  • Create a video: duration adjusts by the second by sliding between chevrons.
  • Create a video: the starting image returns. An imported image becomes the first frame of the clip.
  • Create a video in Ultra: the voice recorded in a character's card gives them their voice in the clip, no extra charge. The speaker icon on their card mutes it.
  • Create a video: the AI write button works for the chosen engine, announces the language of each line and no longer invents dialogue. A character you don't have speak stays silent even if asked a question.
  • Online video: audio from Ultra clips is leveled from one clip to the next, and voice is better synced to lips.
  • Online video: an Ultra clip already computing doesn't cancel anymore, it's delivered and charged. The Cancel button stays clickable and only says so if you click too late.
  • Starter: price shows before the image. The clip goes in 480p, 720p if you request HD, 1080p if you write it, and the written duration (4 to 30 s) is rendered and charged as is.
  • API and MCP: Ultra and Multi-Ref Ultra at per-second pricing (480p, 720p or 1080p), like in the app.

v0.8.7

  • Chat: in a photo where you are with the character, your position is carried over (lying down, sitting, standing), and no one keeps their shoes on the bed. When you lead the action, you move to the foreground of the image; when you only watch, you stay in the background.
  • Chat: when you write or call each other, the character knows you are not with them, and photos no longer show you next to them. You join them when the conversation tells of your arrival.
  • Chat: the character no longer appears doubled in a photo or video where you are with them. Underwear stays in the prompt when visible, even with a shirt bottom or pants alone.
  • Chat: photos and clips of Lucas show his age, 24.
  • Chat: when auto photo or video is on, its price shows next to the send arrow, like in Create.
  • Chat: the percentage of a film follows its true length, instead of stopping around halfway when the video arrives. Toward the end, the bubble indicates the clip is coming.
  • Transition, video continuation: the note fits on one line, under Format and Resolution ("Video continuation from Start, sound included. Forced size: …"). With a continuation placed, the screen no longer scrolls.
  • Generate: locally, a character without LoRA created by its characteristics alone generates instead of displaying "character without identity".
  • Multi-Ref Image: a copied image pastes with Ctrl+V and becomes a reference card, without going through Import. Text pasted in the prompt is not affected.
  • Multi-Ref Image: you can generate from text alone, without any cards. With a single character, or without a character, rendering also happens locally on Mac.
  • Multi-Ref Image: when you modify the prompt after generating, the image stays displayed small to the left of the text. Clicking on it reopens it large.
  • Local / Cloud GPU Switch: it always displays the mode actually being used. If you chose to do everything online at setup, it is on Cloud GPU; if you installed the local engine, it is on Local.
  • Multi-Ref Image: on Local, no render goes online anymore, even if you never touched the switch. When your card cannot render the scene, a warning says so and suggests switching to Cloud GPU.
  • Generate and Retoucher: on Local, no render goes online on your behalf anymore, even without a model installed or on a machine without a capable card. A warning says why and suggests switching to Cloud GPU. When a retouch exceeds what your card can do along the way, the "Switch to Cloud GPU" button toggles the switch.
  • Train an image LoRA: on Local, the "online" buttons (complete, train, remake an image) are grayed out and say why on hover; those for your card stay the same. When your machine cannot do the step, a warning suggests switching to Cloud GPU.
  • On Local, when your card's engine is not ready, proposed images in the add window (Multi-Ref Image, Transitions), a LoRA preview, and character views or portrait no longer go online without saying: a warning suggests "Launch online", with its price, for this image only.
  • Create: on Local, nothing goes online on your behalf anymore. A video, attached image, or image without a ready local engine displays a warning suggesting to switch to Cloud GPU. The price of a send now shows next to the arrow.
  • Chat: the Local / Cloud GPU switch no longer appears at the top of the Chat screen.
  • Create: the text input area takes up three lines and what overflows scrolls inside. The paperclip and arrow have their row below the text, which never goes underneath anymore, even on phone.
  • Create: the input fits on two lines and the price shows next to the send arrow. A camera button, next to the paperclip, takes a photo and attaches it as an image. The app opens on Create.
  • Characters: technical files from Face swap and Multi-Ref Video no longer appear as LoRAs to sort ("Type?"), and can no longer be deleted by mistake by attaching them to a character.
  • Create: the paperclip opens "My Creations", your most recent images, the one you just made first. A touch attaches it for retouching, without searching your files, which stay as a button below. The message "Two images maximum" clears as soon as you remove one.
  • Characters: the profile photo changes again, with or without LoRA. Clicking the photo on the card opens "My Creations" or your files. Without a character sheet, this photo plays the character in Multi-Ref and Face swap.
  • Create: if the app restarts during an online render (screen off, another app open), it resumes tracking on return, and the image or video arrives in the feed and Gallery instead of being lost.
  • Create: screen and startup assistant texts take their accents in French ("What do you want to create?").
  • Each image and video created carries in its metadata the mention "generated by AI" (IPTC field read by major platforms), as required by European AI regulation. Invisible to the eye: the image and video do not change. Photos imported to train a character do not receive it.
  • Multi-Ref Video: in Ultra, a reference strength different from 1 works again (rendering failed after several minutes of waiting).
  • Render queue: each render in progress displays its elapsed time next to the percentage ("3:12 · 84%"), on your card as well as online.
  • Face swap: click a card to adjust it directly in the generator, without a window: its character, its voice (the one from the character's card, at the same pace as yours, or your original voice), and its scene, what the character does during their passage ("@Kara puts on headphones"). The @ mention of the character carries its description: click on it to see it and correct it.
  • Face swap: the character has more space around you, their ears, wings, or large frame are no longer cut off. Two adjustments per card, in "Advanced" above the scene: the character's area (Normal, Large for big ears and wings, or Tight to keep them close to you) and resemblance, which adjusts how much they stick to their card.
  • Face swap: "Retry" in the gallery reopens a Face swap with its video and all its cards, ready to modify or relaunch.
  • Face swap: open to free trial. Without a license, on Cloud GPU, a 5-second card uses the day's trial video; on your graphics card, the trial renders freely.
  • Face swap: rendering is faster. The area around the character calculates in seconds instead of several minutes.
  • Face swap: before launching, the app checks that your balance covers the displayed price, and suggests recharging if not. A render of more than 1,000 credits asks for confirmation, with its price and computing time.
  • Face swap: a video preview with a playhead, to align the start and end of a card frame by frame; card edges snap to it. The spacebar plays and stops.
  • Face swap: the character's description (male, female, creature) starts from their card and corrects on each card. A video dragged from the file explorer places directly, and an interrupted render (app closed) resumes without recalculating everything.
  • Subscription: a subscriber's license is read every 15 minutes and when returning to the app. An ended subscription switches them back to demo mode, offline too once the paid period is past, with a screen that explains it and opens the tendre.ai account. A renewal makes it active without restarting.
  • Purchase screen: the subscription at 19.90 € per month, with 200 cloud credits each month, replaces the Personal lifetime license, next to the Pro license. A subscriber switching to Pro does so from their tendre.ai account, at full price.

v0.8.6

  • Windows: the app is now called tendre.exe (Task Manager). The update removes the old dcrmer.exe and keeps your Desktop and Start menu shortcuts.
  • Windows: installing in a protected folder like Program Files now asks for administrator rights instead of failing; a refusal is explained in your Windows language. Uninstalling does the same.
  • Video and Multi-Ref Video: the AI button now writes the scene like a director. The main character opens the shot, each object not already in frame enters by an action, a tricky gesture is told over several beats, and detail lands at the key moment. In automatic duration, it writes enough time for your target length.
  • Chat: automatic photo is on from the start. Its button, bottom left of the input, now shows a camera and the word Auto, with the price of one photo next to it while it is on. A click turns it off, and this choice is saved.
  • Chat: when today's trial or credits run out, automatic photo turns itself off until the next time you open chat, with just one warning instead of one with each reply.
  • Chat: the input area is three lines tall and anything extra scrolls inside it. Text stays above the button row (auto photo, send) and never goes underneath, even on phone.
  • Chat: reopening a conversation during the day, the scene picks up at the same place instead of inventing another one.
  • Chat: if the character steps out of role to comment on the request, their reply is redone, and a past slip no longer taints the rest of the conversation.
  • Chat: when she is watching a screen (cinema, television) facing the camera, the photo no longer shows the screen behind her; only its light on faces remains.
  • Chat: a photo of her standing in a room no longer makes you the main subject of the image.
  • Chat: the free trial gives you a real 20 replies per day. The work done behind the scenes with each message (tracking the scene, preparing a photo) no longer uses them up: the trial sometimes stopped after six messages.
  • Chat: videos speak with the voice chosen in the character sheet, the same as in Create a Video and Multi-Ref. A voice set once for chat alone is no longer used.
  • Chat: with a license, Megan's Chat videos use her own video model. Her face holds better when she speaks. Other characters will follow once they have theirs.
  • Chat: Lina's robotic arm is her right arm, like in her images. She says it that way, and her photos no longer ask for a flesh hand on that side.
  • Create a Video: the chosen character is placed in the prompt by their @Name tag, under the field, like in Multi-Ref. They are called only if their @Name is in the text, and clicking it shows what they send to the model and lets you change it: their video LoRA word, their age and type. Their name and character sheet description no longer go in secret at the head of the prompt.
  • Create a Video: all characters are offered, even without a video LoRA. Their @Name then carries their full sheet (type, age, hair, eyes, build).
  • Multi-Ref Video: a character linked by their LoRA is called only if their @Name is in the text; otherwise, a note says so under the field.
  • Create a Video: the AI button keeps the character's @Name when rewriting the scene, without describing their looks, which the @ already brings.
  • Multi-Ref Video: a character speaks with the voice from their sheet, online or local. The model renders it itself, with their lips and scene sounds. Two voices at most, for the first two characters linked by their images.
  • API Pro and MCP: in Multi-Ref Video, the first two characters can speak with their own voice (`voice_b64` in the API, `voice_url` in the MCP), like in the app.
  • Multi-Ref Video: a voice icon on each card says whether the character's voice is going out, and clicking it turns it off or back on, for a character who does not speak. Beyond two voices, or sound off, it is greyed out and a bubble explains why.
  • Create a Video, Multi-Ref Video and Transition: a Framerate button picks 24, 30 or 60 frames per second, 30 by default, for sharper motion. 60 is most fluid. Price does not change, but maximum duration follows frame count: at 60 fps, a clip runs up to 8 s, 12 s in enhanced consistency. The next clip keeps its framerate. With sound too. Locally too, sound included.
  • Create a Video and Multi-Ref Video: in Quality at 30 fps, a clip runs up to 20 s in fixed duration (16 s before), and in Ultra too in Create a Video, up to 1080p. At 60 fps, in Quality in Create a Video: 16 s with sound, 20 s without (8 s before). Beyond 16 s, the format price doubles, like for long clips in enhanced consistency.
  • API Pro and MCP: at 30 fps, `length` runs up to 601 (20 s) in Quality, Multi-Ref included, and in Ultra up to 1080p (Ultra fixed price doubled beyond 481 frames). At 60 fps, in Quality: 961 with sound (16 s), 1201 without (20 s), at double price beyond 481. Beyond 481 frames, in fixed duration, the format price doubles (long tier, as in `consistency` `enhanced`).
  • Transition: in enhanced consistency at 60 fps, a shot runs up to 16 s (12 s before), at the price of long clips; with sound and keyframes or an end image, it stays at 12 s.
  • API Pro and MCP: `end_image_strength` sets the firmness of a transition's end, from 0.3 to 1: 0.4 loose, 0.75 normal, 1 firm, the three choices in the Transition app, in standard consistency or enhanced.
  • Transition: in enhanced consistency, the firmness of the chosen end (Loose, Normal, Firm) is respected; the end always came out firm.
  • A video played fullscreen displays in full: a portrait clip is no longer reframed as landscape to fill the screen (Transition, gallery and other previews).
  • Transition: the start can be a video from your gallery. At Start, the Videos tab shows your clips (played on hover); the Transition picks up the motion from their last second, then goes through keyframes and the end image. Format, resolution and framerate are the clip's, so the splice does not show: the clip you get starts where the other ends, just put them back-to-back in Edit. Clips up to 1080p.
  • API Pro and MCP: in enhanced consistency, a clip asked for in 1080 wide comes out in 1088 x 1920, like in Quality (it came out in 1024 x 1920), and 720p in 768 x 1280 (704 before).
  • Video LoRA, step His Voice: Use the voice from the clips moves first and becomes the default choice. A character who already has a voice in their sheet keeps it.
  • Character sheet, associated LoRAs: the folder icon appears only if the file is on your computer. Otherwise, the line says Online only or Not on this computer. On Mac, the icon showed a red error (file not found) for a video LoRA trained online.
  • Video LoRA: a character's file trained online is now downloaded to your computer, on Windows and Mac alike, even without a video engine installed. A LoRA already trained that was only online comes down when you open their sheet, which shows progress. If a video engine installs later, it picks up the file without downloading it again.
  • Image LoRA: same rule. A character's file trained online, or received by a share code, is downloaded to your computer, on Windows and Mac alike, even without an image engine installed. A LoRA already trained that was only online comes down when you open their sheet. If the image engine installs later, it picks up the file without downloading it again.
  • Video and Multi-Ref Video: a character whose sheet has only their traits (age, hair, eyes, build, etc.) carries them in their @mention, like in Generate. In Multi-Ref Video, a character without a character sheet, played by their photo alone, sends their full description; with a character sheet, nothing changes.
  • Transition, next in a video: the Start card says Video instead of Import. Under resolution, the note now says the clip's sound continues (sound included), in Transition as for Next Shot in Video; for a silent clip, it still says sound starts fresh.
  • Video LoRA, step Their Clips: with clips brought in, the grid, the note to make and remaining clips count only what is left to make, like the button. With 27 clips brought for a grid of 36, the screen showed 63 boxes and 36 to make instead of 9. French text in steps Their Voice, Their Clips and Training gets accents back.
  • Transition (next in a video) and Next Shot: sound continues at the splice. The next one picks up the sound from the end of the clip instead of making a new one, which could start with a second of silence.
  • Video and Multi-Ref Video in Ultra: sound set to Mute gives a truly silent clip. The clip used to come out with made-up sound and voice.
  • Free trial online: the voice goes right with the clip, the one you picked in Video or the character's voice in Multi-Ref. It was dropped with no word about it.
  • Prices shown before a video render come from the server grid, the same one that bills: a price cannot differ from one screen to another. With no server connection, no price shows.
  • Multi-Ref Video: sound is picked in options, Mute or With Sound, next to framerate.
  • Create a Video and Transition: a Quality 480p video without sound costs 2 credits, any duration (5 before). Multi-Ref and enhanced consistency keep their price.
  • API Pro and MCP: a Quality 480p video without sound costs 2 credits (Multi-Ref and enhanced consistency: 5).
  • Create a Video, Multi-Ref Video and Transition: a video online without sound comes out truly silent. The engine used to add a sound track still.
  • API Pro and MCP: a video asked for with no `audio`, or with `audio` false, comes out truly silent, like the docs say. The engine used to add a sound track still. Ultra keeps sound by default.
  • Create a Video and Multi-Ref Video: Fast mode is removed. A video is made in Quality or Ultra. A Fast clip taken from the gallery starts over in Quality.
  • API Pro and MCP: the `turbo` mode is removed. A `turbo` request is rendered in Quality 480p, same orientation and length, at Quality 480p price (2 credits mute, 10 with sound).
  • API Pro and MCP: field `fps` (24, 30 or 60, and 50 in enhanced consistency for a slight slowdown), 30 by default, with or without sound (50 without sound only). `length` counts at this rate: with no `fps`, a clip of the same `length` runs shorter than before (241 frames: 8 s instead of 10 s); `fps: 24` keeps the old behavior.
  • Generate and Retoucher: the chosen character is placed in the prompt by their @Name tag, under the field, like in Multi-Ref. They are called only if their @Name is in the text, and clicking it shows what they send to the model (their LoRA word, their age and type, or their description) and lets you change it. The LoRA word no longer goes in secret at the head of the Retoucher instruction.
  • Generate and Retoucher: the AI button keeps the character's @Name when rewriting the prompt, without describing their looks, which the @ brings already. It goes through the local assistant when installed.
  • Multi-Ref Image: the @Name of a character played by their LoRA goes by their LoRA word, with no name in front (linab3yndai dancing at club, not Lina, linab3yndai, dancing at club).
  • Character sheet: a pencil next to Redo shows the text sent to the model for this view and lets you adapt it when the result does not fit. The changed text is kept for the view and goes with each Redo; Automatic Text goes back to the app's.
  • Generate, Retoucher and Multi-Ref Image: an @Name typed without capitals (@lina for @Lina) calls the character right. It colors and opens like the others; before, it went to the model as typed. In Multi-Ref Image, @Lina no longer bites into the @ of another card starting the same way (@Linabelle).
  • Multi-Ref Image locally: the message sending you online at last says what the machine can do (up to four reference images, trained characters, and both mixed). It no longer shows with no card, or on top of the mix message.
  • Multi-Ref Image: a character without a character sheet is no longer said to be ignored by their LoRA, since they can be played by it. A card's strength explains what it does, and that of a character played by their LoRA caps at 1.5, the value the render truly gets.
  • Generate: a character without a LoRA is offered as soon as they have traits, and their @ carries their description (type, age, hair, eyes, height, build, marks). They were missing when their description was never saved, like characters from the API, who no longer have a made-up age either.
  • Multi-Ref Image: a character without a character sheet, played by their profile photo, carries their full description, not just age and type.
  • Generate and Multi-Ref Image: picking a character, or adding their card, writes their @ right into the text. No more need to click their tag under the field, which stays there to place them elsewhere.
  • Multi-Ref Image: a character made only with their traits, no image or LoRA, now generates. They enter the scene by their description, alone or with other characters, objects and locations. The list says Description for this character, not Photo.
  • Generate and Multi-Ref Image: the AI button accounts for Their World on the sheet of called characters (era and place, job, wardrobe, what they never wear), for what your idea does not spell out. What you write always goes first.
  • Generate, Retoucher, Multi-Ref Image, the Starter, character sheet and LoRA training: shown prices come from the server grid, the one that bills. With no server connection, no price shows, and the window offering to go online no longer gives a made-up amount.
  • Character sheet: each LoRA shows its own keyword, one for image and one for video. The word of a LoRA trained in the app does not change: it is the one from training. That of an imported LoRA can be fixed and may have spaces, and changing it no longer touches the character's other LoRAs.
  • Settings, Generation: a block Engine Memory (local) shows the memory Windows can still set aside. When it runs out, the engine fails even if free RAM is left. The Clear Engine Memory button restarts it so it gives back what it took; it waits for renders to finish. On Windows, a second button opens the virtual memory window, with steps to make it bigger.
  • Local image engine on Windows: a character sheet retouche could run Windows memory dry and crash the app with it. The engine now sets aside half the memory for the same image, identical.
  • Character sheet: full-body front and back views show a human character, man or woman, arms slightly out from the body, not stuck along it anymore. A creature stays as before.
  • Characters: deleting a character now asks for confirmation.
  • Chat: unlocking a paid character again shows its confirmation request with the price; it was not showing, and the unlock was not going through.
  • Objects and animals, Locations and decors, API keys: deletion really waits for your answer. The question showed but erase went without waiting.
  • Character sheet: the third view fits a three-quarter bust or a side full-body, turned the way you pick, and the fifth view is the back. Generate always renders the three-quarter bust.
  • Image AI assistant (Generate, Retoucher, Multi-Ref Image): a character called by their @ is no longer followed by their copied description (@ Lina, a 30 year old robotic-woman, stands becomes @ Lina stands). An action detail stays, and a non-human character keeps what they are (a baby dragon).
  • Generate, Retoucher, Multi-Ref Image: the chosen style is added at the end of the prompt without leaving a comma after a period (dancing., photorealistic becomes dancing, photorealistic).
  • Mac: videos go fullscreen with their player button (gallery, chat, Video, Multi-Ref, Edit), like on Windows.
  • Gallery, videos: the circle picking a video moves to top right of its thumbnail. On Mac, the player buttons hid it.
  • Gallery, images: the pick circle moves to top right too, so both galleries match.
  • Gallery: Select all and None are in blue. Red stays for actions that remove something.
  • Confirmations (delete a character, object or place, revoke an API key, unlock a chat character) open in an app window, in its style, not the grey system box, which could open behind the app and block it. Esc or Cancel refuses, and the Cancel button has focus.

v0.8.5

  • Menu: the New thumbnail switches from Multi-Ref Video to Transition.
  • Multi-Ref Video and keyframe window: clicking an @mention shows what it sends to the model and lets you edit it, just like in Multi-Ref Image; nothing is added behind the scenes anymore.
  • Multi-Ref Video and keyframe window: deleting multiple @mentions added by buttons also clears the colored boxes (deleted text could still reach rendering).
  • Transition online: while waiting for a machine, the screen says so (Waiting for an online machine, startup may take a few minutes) instead of a frozen bar at 25%.
  • Transition, generated keyframe: an @Name label below the field places the chosen character where you want in the sentence, like in Generate; changing character updates the mention.
  • Multi-Ref Video: changing a card's reference updates its @mentions in the scene, like in Multi-Ref Image.
  • Multi-Ref Video: Style and Framing menus above the scene, like in Multi-Ref Image.
  • Transition: with enhanced coherence, the clip has your chosen duration (it used to run on auto duration, and an end image made rendering fail).
  • Multi-Ref Video: a character with a video LoRA can run via that LoRA instead of images, your choice in Add a subject like in Multi-Ref Image, locally or online; one LoRA per clip. The dot on the card reopens the choice.
  • Gallery: Resume on a clip made with Transition reopens Transition, not Generate a video: text, duration, format, resolution, sound, coherence and keyframe positions come back; images need repositioning.
  • Multi-Ref Video: the app's official characters no longer appear in character selection; only your own characters (and storefront ones) are offered, like in My Characters.
  • Multi-Ref Video and AI video assistant: a line written in quotes is no longer translated; a line in French is spoken in French in the video.
  • Multi-Ref Video: after rendering, the preview fits in the window without scrolling; when choosing a character sheet, the hovered card gets a blue border, like in Multi-Ref Image.
  • Transition: a timeline replaces image fields. Start and end images optional, up to 4 keyframes placed on the timeline at the moment you want, each with its strength; the video passes through each one. Online in Quality mode like locally.
  • Transition: Enhanced Coherence option (more stable throughout the clip), online; it allows clips up to 30 s and turns on by itself beyond 20 s (15 s with an end image). A clip from 21 to 30 s costs twice the price of its format. Also open to free trial, up to 20 s; from 21 to 30 s, reserved for licenses.
  • Transition: new screen. Timeline and generator form one scene; a window lets you pick each keyframe from the gallery, import it, or generate it with a character by LoRA. During rendering, the timeline folds and the video displays in its place, at true size.
  • Interface: screen name no longer repeats below the top bar, and Generate, Retoucher and Transition fit in the window without scrolling. An (i) next to a setting explains what it does.
  • Multi-Ref Image: a new tab to compose an image with multiple characters, objects and a location. Each character plays via its LoRA for truest likeness, or by an image picked from its character sheet. On Mac, rendering happens online.
  • Multi-Ref Image: without LoRA, reference images (people, objects, location) are truly captured, like attached images in Starter, online and locally on PC, 2K included. Beyond two images, each extra person is added by one more retouch: rendering takes longer and costs one image per retouch, price shown on the button.
  • Multi-Ref Image with LoRAs: price follows character count, one image for one character, then one more per character from two on (3 for two characters). Location doesn't count.
  • Multi-Ref Image: a character via LoRA and another via an image from their character sheet can now be in the same render. Price counts each step (composition, LoRA pass, image pickup).
  • Gallery: Resume an image made with Retoucher reopens it in Retoucher, with its prompt and starting photo, instead of Generate. Retouches made before this version come back with their prompt only.
  • Retoucher: two modes. One Image retouches the photo with a prompt (brush zone, element pasted by hand); Two Images runs Starter mode: the second image is redrawn in the scene per your prompt, locally or online (online on Mac). An Image Weight slider gives more space to image 1 or image 2. Brush size opens under its icon, and the field example shows how to name image 1 and image 2. Ctrl+V pastes an image: photo first, then second image. Each image has its thumbnail and X.
  • Retoucher: Style menu like in Generate; chosen style adds to the prompt and photo fidelity drops to 2, else it won't work.
  • Retoucher locally: a photo too large for your GPU offers, before rendering, to shrink it to 2048 px, go online, or cancel.
  • Generate: what menus add (character, style, LoRA) writes into the prompt field, where you can move or edit it; text runs as displayed. If a LoRA's trigger word is missing from the text, a warning flags it.
  • Multi-Ref Image: an LoRA Pass slider on each character rendered by LoRA sets how much its LoRA repaints its zone.
  • Local engine: Cancel really stops rendering on your GPU, a long render is no longer abandoned while the engine works on it, and an unresponsive engine is shut down with a clear message. Engine errors are translated.
  • Free trial: creating a character is open; LoRA training stays for licenses only. The High tier (2K) locked in Generate leads to pricing.
  • Chat: a stage direction written in Dutch or German translates correctly before going into the photo.
  • First install: a Do you have a license? step before picking features. Activate right away (license file or activation link) also opens reserved features, like Uncensored Mode video; otherwise, continue in demo mode.
  • Mac: at install, a folder the app can't write to explains the macOS issue (the message talked about Windows and Program Files).
  • Install: model download follows one progress bar, by weight, with amount downloaded over total (8.4 GB of 42.6 GB), not a file count.
  • Mac: right after local engine install, the Local / Cloud GPU switch and Zoom are there; you had to restart the app to see them.
  • Chat: the character says from their first line whether you're with them or not; when you write from far away, you no longer show in their photos, and a simple invitation to come no longer makes you appear next to them.

v0.8.4

  • First installation: below the download bar, the amount already downloaded and the speed ("7.5 GB of 13.1 GB · 90 MB/s"); on Mac, the speed.
  • Settings > Components: a model that installation downloads in the background shows "Downloading..." (or "Waiting" if it hasn't started yet) and its button is disabled, so it can't be downloaded a second time by clicking.
  • First installation: a shorter screen, one sentence per feature, accents restored.
  • Chat and Gallery: videos and clips received in chat appear in Gallery > Videos, with their thumbnail, and play in full there, as in the conversation. A video that arrives incomplete is re-downloaded.
  • Training: retraining a character deletes the old version of its LoRA; removing a LoRA from a character sheet also deletes its online copy.
  • Settings > Components: the "Refresh" and "Import a model" buttons disappear. An imported model wasn't loaded by any feature; each one now installs the models it chose and tested. Sections have feature names (Engine, Image Generation, Video with Sound...); model names stay on their lines. Models that Tendre.AI can no longer use are grouped under "Legacy Models", with the space to recover and a button to delete everything. A download now shows its progress (percentage, speed, Cancel button) on its line and in a bar at the bottom of the menu, and continues if you leave Settings.
  • Settings > Components: after installing a video component, the local video engine restarts on its own to use it, without interrupting a render in progress.
  • Video AI assistant (Create Video, Multi-Ref Video): when the machine's assistant rejects a scene, the rewrite goes online; a rejection no longer overwrites the text, which comes back unchanged with a clear message.
  • Generate: in High quality (2K) with an online render, the displayed resolution is that of the output image (1184 × 2048 in portrait, not 1536 × 2688).
  • Generate: the screen fits in the window. The preview takes up all available height without extending below, and the form scrolls on its own when it's long.
  • Retoucher: the photo and its tools fit in the window, without extending below, whatever the length of the help displayed under it.
  • Generate and Settings: the 768 quality tier is called "Small (768)" instead of "Small PC (768)".
  • Online Library: objects, animals, and locations follow the license from one machine to another and in the API; description limited to 300 characters, images at least 512 px.
  • Multi-Ref Video: in the window of an object, animal, or location, the X that removes an image moves to the top right.
  • Multi-Ref Video: the scene field expands by dragging the handle at its corner, on Mac too (it wasn't responding).
  • Character sheet: the character's description (distinctive marks, prosthetics, tattoos...) and age go with its LoRA in the generated views.
  • Local engine: modules from the old image engine, unused since 0.6, are removed on startup; the engine starts leaner.
  • Settings > License: a subtle link opens the list of software and models from other authors used by Tendre.AI, with their full licenses.
  • The "What's New in This Version" window returns on first launch after an update, with the 0.8 features (it hadn't shown since 0.7). It displays only once for all of 0.8.
  • Mac: installation no longer says retouching stays online; it runs fine on Mac.
  • Chat: when a photo shows the character with you without naming you (two riders, gazing at each other), you're drawn properly, no more duplicate character.
  • Chat: real time is only used at the start of a scene (conversation reset, or 2 hours without a message). After that, the story moves at its own pace and the character remembers the moment and day it has reached, until early morning, even when you reopen the conversation.
  • Chat: the video requested from the character is now 10 seconds, in 480p with sound, like "Transform into Video", and costs 20 credits.
  • Chat: asking the character for a video works in all phrasings and in all nine languages ("send me a video", "film yourself", "a video of you"...), and they reply with a video, not a photo. A simple mention of a video no longer triggers anything.
  • Chat: two-person photos more faithful. You can appear in the image, face included; your description (including hair) is no longer erased or placed twice; the scene is no longer cut off mid-word; a character's style (cyberpunk) no longer adds rain and neon when the scene has its own location.
  • Chat: a paid unlocked character shows their face from the first photo, without waiting to restart the app.
  • Starter: an attached image really changes style when the instruction asks (realistic photo, anime, painting...); before, retouching kept the trait of the original image.
  • Free trial: when the day's trial is used up, the message says it comes back tomorrow.
  • Starter: videos in the feed play (the Play button had no effect).
  • Starter: the image or video received fits in the feed, without extending in height.
  • Starter: a duration written in the request ("10 seconds", "a 20-second clip") sets the video duration: 5, 10, 15, or 20 seconds.

v0.8.3

  • Multi-Ref Video: in the "Add a subject" window, names display in full, the "Character sheet" or "No character sheet" label moves below, and the "Create a character" tile stays visible even with many characters.
  • Multi-Ref Video: a long scene text, especially from the AI assistant, no longer passes under the @ tags in the field.
  • Multi-Ref Video: in Local mode, the message offering to download the Multi-Ref component (or explaining what is missing) appears in blue, like the one in the Video tab.
  • Multi-Ref Video AI assistant: it no longer automatically places characters on the left or right of the image; their placement follows what the scene says.
  • Multi-Ref Video: in "Add a subject" and "Choose a scene", hovering over a tile shows a pencil that opens the character sheet, or the window for the animal, object, or location to edit it without leaving Multi-Ref.
  • Multi-Ref Video: the old "Pool house" scene, leftover from the SDXL era, disappears from the scene list.
  • Chat: a response where the character breaks character to refuse, like an assistant would, is now better detected and rewritten, and it no longer kill the mood of the conversation.
  • Chat: the price of a photo or clip displays with "=" instead of "≈", it is the exact price.
  • Windows: GPU training now finally uses imported photos (before, only the generated sheet went there, and without a sheet the "empty dataset" error appeared).
  • Training: 40 images maximum, no duplicates, and the screen indicates the excess ("74 · 40 used").
  • Training: photos drop directly onto the grid, several at a time.
  • Characters: the profile photo created after training respects the character's gender and age, and regenerates with one click on the sheet.
  • Characters: each LoRA attached to a character can be removed from its sheet (the file is kept).
  • Training: a green notice signals the end of training; clicking it opens the character sheet.
  • Training: the remaining time starts from the estimate announced at launch, then follows the actual pace (it sometimes announced an hour for a few minutes).
  • Training: retraining a character gives the new face everywhere (the previous one is no longer served by mistake).

v0.8.2

  • Starter: a second image can be attached (two maximum): the person or outfit to put in the first one, according to the description.
  • Starter: an attached image prepares faster and sends lighter, and the wait icon spins during its import, including pasting.
  • Starter: "Resume" in the Gallery puts the image or video text back in the Starter.
  • Starter: "Resume" also restores the attached images from the creation.
  • Starter: switching tabs during a creation no longer makes it disappear. On return, the bar, timer, and Cancel button are still there, and the result displays as soon as it is ready.
  • Online images: the progress bar advances throughout the calculation instead of jumping all at once at the end, and the Starter shows a timer next to it.
  • Generation: common English words (gaming, analyzing, cocktail...) no longer trigger the "Forbidden content" rejection by mistake.
  • Translations: a few labels in Spanish, Portuguese, and Polish get their accents back.
  • Starter: the X on an image or video removes it from the discussion, without deleting it from the Gallery.
  • Gallery and generation tab: a long press on an image offers to copy it or save it to the phone, and on a video in the Gallery to save it. Saving drops it in the Tendre.AI album, which the phone's Gallery and other apps see (Android 10 or newer).
  • Gallery: "Copy" in the selection bar copies the selected image.

v0.8.1

  • Multi-Ref Video: a character's likeness comes from their character sheet (5 views), not their LoRA. A character without one is flagged on their card ("No character sheet", click to open their profile), and a warning reminds you when generating, with the choice to create the sheet or generate anyway.
  • AI Assistant: a character who is not human (a creature in Characters) is no longer rewritten as a person in the video text.
  • Windows: when the local video engine is slow to start, a "Show engine log" link opens its log, to send to support.
  • Menu: the "New" thumbnail for Multi-Ref Video no longer overflow the sidebar, and the name displays in full.
  • Starter: the X on an image or video removes it from the conversation, without deleting it from the Gallery.

v0.8.0

  • Multi-Ref Video (NEW): a video from references. Up to four subjects (characters, animals, objects) and a scene connect to the generator, and the text calls them by name (@Lea, @Lucas). Each character is rendered from the views on its character sheet.
  • Multi-Ref Video: Quality and Ultra modes, portrait, landscape and square formats, in 480p (Quality), 720p or 1080p, automatic duration or 5, 10 and 20 s, with or without sound. Price is that of a clip in the same format; Ultra costs 40 credits in 720p and 80 in 1080p.
  • Multi-Ref Video: a strength per reference (0 to 2) sets how much space it takes in the clip, and a seed locks in to redo the same shot.
  • Multi-Ref Video: the progress bar tracks the actual render duration, and signals waiting for a machine then clip finalization.
  • Library: new pages "Objects and Animals" and "Locations and Scenes". Each item gathers up to five images, a title and a description, and connects into Multi-Ref Video.
  • Gallery: "Resume" on a Multi-Ref clip reopens Multi-Ref Video with its references reconnected, its text, format and seed.
  • Demo: Multi-Ref Video can be tried with the free video, in 720p and up to 5 s.
  • Windows: Multi-Ref Video also renders on your machine, in Quality mode, with a capable card. The required component (2.2 GB) downloads from the tab or from Settings > Components.
  • AI Assistant: the AI button on Multi-Ref Video and the Video tab rewrites the idea into text ready for video (shot, action, camera, light, sound, dialogue), keeping the @references.
  • AI Assistant: it installs on your machine (2.8 GB), on first click on AI or from Settings > Components, and rewrites then offline, without going through our servers: on the processor on Windows, on the Apple Silicon chip on Macs.
  • AI Assistant: without installation, rewriting happens online, free; in demo, it counts as a trial chat message.
  • Video: an online render is no longer abandoned after 20 minutes. It waits up to 30 minutes for a free machine, then has 30 minutes of compute; if abandoned, it is cancelled and nothing is charged.
  • Mac: when Tendre.AI is launched from the disk image, the update explains it must be dragged to Applications, instead of failing silently.
  • Settings: an update installation failure no longer displays "Check impossible".
  • Characters: new "My Characters" screen. Each character has a unique card in four editable sections in any order: traits, character sheets, voice and LoRA training. Everything saves automatically.
  • Characters: the eye color chosen on the card now enters the character description used for their images.
  • Characters: with a license, character sheet views are copied online, encrypted, and found when opening a card if a file has gone missing from your machine. The copy follows the card: a view, sheet or character deleted is also deleted online.
  • Characters: Lea and Lucas arrive with their complete character sheets (five views), visible on their card, even in demo.
  • Characters: Lea and Lucas cards are read-only: you can view or delete them, not edit them.
  • Character sheets: the front view generates better from a three-quarter photo, and the real photo serves as the priority model.
  • Characters: a character can be something other than a woman or man ("Other...": a dragon, a robot, an elf), and their age can stay unspecified. Their character sheets adapt, for example without shoes for a creature.
  • Character sheets: a view being generated can be cancelled, and generated views are saved in the image folder chosen in Settings.
  • Character sheets: an imported photo on a plain background is auto-cropped to its subject.
  • Character sheets: a remade or deleted view goes to system trash instead of being erased.
  • Training: a Cancel button exits the assistant and returns to the character card, without losing photos already added.
  • Interface: a banner at the top of each page gathers the page name, Local / Cloud GPU choice, credits and Refresh button. It no longer covers page buttons.
  • Interface: new sidebar, which shrinks to its icons with one click (Shrink button at bottom) to leave more room for pages.
  • Installation: a network cut of a few minutes no longer interrupts a download, and a stopped download (longer cut, app closed, PC restarted) resumes where it stopped when you restart it, instead of starting over.
  • Windows: if Tendre.AI is still open during installation, the installer asks you to close it instead of showing an error.
  • Starter: a copied image pastes directly in the field (Ctrl+V, or Paste on phone) and becomes the attached image, like with the paperclip.
  • Starter: the crosses to remove an image or result are properly centered.
  • Chat: the character forever remembers what they learn from you (habits, passions, tastes, work, family, sorrows), even after "Clear Conversation". "Clear Memory" erases it.
  • Chat: an arriving clip plays by itself, muted if the system blocks sound.
  • Chat: fewer duplicate characters in photos, when the same sentence says one standing and kneeling or when you are only referred to by "he".
  • Chat: Megan's hair color is corrected in photos.
  • Translation: "dress" translates properly to a dress and not a robe, and text already in English over several lines is no longer retranslated.
  • Installation: the Conversation line displays "online" instead of "0 GB".

v0.7.2

  • Retoucher: a slider adjusts the number of calculation steps (6 by default, from 4 to 10), on your machine and online. At 6, the result is the same as at 10, almost twice as fast.
  • Mac: retouching now happens on Mac, locally, on the same Macs as local generation. The local engine updates itself on launch (about 2.7 GB to download), keeping models already installed.
  • Chat: the character remembers the current scene better: the location, what you're wearing and where you are.
  • Chat: "Redo" no longer repeats the character's identity in the photo.
  • Chat: Megan looks more like herself in chat.
  • Chat: when you describe your gesture, the photo no longer gets the wrong person ("l'" was sometimes translated masculine).
  • Chat: "Redo" a photo remakes it in the scene as it was at that moment.
  • Characters: the import button is called "Import a character".
  • Characters: importing and exporting a character also works without a local engine, to use it on Cloud GPU. At the first click on Generate, the app offers to send your LoRA online.
  • Characters: a character whose LoRA is already online exports without the file, much lighter. On import, its image LoRA and video LoRA are copied to the online space of the account that receives it, and its voice comes with it.

v0.7.1

  • Logo: the app logo takes up the one from the site, its two colors and its typeface.
  • Mac: video thumbnails display in the gallery and editing (they stayed black).
  • Mac: the gallery selection bar displays in full from the first checked video.
  • Gallery: you leave selection mode with Escape or a click in empty space; the Cancel button moves to the head of the bar.
  • Mac: Retoucher displays like the other entries; in local mode, the page says retouching is done online for now. Settings no longer display a video engine folder, video moving online on Mac.
  • Settings: a wrong folder path could display for the video engine; it is fixed on launch.
  • Editing: you choose where to save the exported file, then a button opens its folder.
  • Editing: MP4 export works without installing anything. FFmpeg is now shipped with the app, on Windows and Mac, and also serves voice on video and import of personal clips.

v0.7.0

  • Mac: Tendre.AI arrives on Mac with Apple silicon (M1 and newer, macOS 13 minimum). The app is signed and notarized by Apple, and updates itself automatically like on Windows. Everything works online on any Mac.
  • Mac: local image generation, with characters and enlargement, on Macs from M4 with 24 GB of memory, or M2 or M3 Pro, Max or Ultra with 24 GB. The installation assistant offers the local engine on these Macs and checks online rendering by default.
  • Mac: retouching and video are done online for now.
  • Settings: the Components page (the old Models tab) is now found in Settings.
  • Demo: the Local/Cloud GPU switch also appears in demo mode, on machines able to generate locally.
  • Generate: on Cloud GPU, no more offer to send Lea or Lucas online, already available on our servers.
  • Chat: a photo or video received now displays in full, the thread goes down to its bottom once it loads (without bringing down a conversation you scrolled up to reread).
  • Training a character (image): the number of steps adjusts again, from 250 to 2000, set to 1000. 250 is enough for a quick test. The online price follows the slider (60 credits for 1000 steps, from 21 to 112), as does the stated duration, online and on your graphics card.
  • Training: character training appears in the render queue, with its progress and a Cancel button, online and on your graphics card.
  • Training: canceling during photo upload really stops online training, it no longer goes through anyway.
  • Training: photos added to a character stay in place after restarting the app, and a removed photo no longer returns.
  • Characters: a character's preview follows the Local/Cloud GPU switch; without a graphics card, it renders online (1 credit, like any image).
  • Characters: Regenerate preview works again for a character that keeps an old engine LoRA alongside its Krea 2 LoRA.
  • Characters: an X removes an old engine LoRA from the card, after confirmation; the file stays on disk.

v0.6.9

  • Video: after passing through Fast mode, a Transition or a clip taken from the gallery could go on an old engine, with no mode selected on screen (and in Transition, with the Fast mode price displayed). It now goes in Quality, at the displayed price.
  • Video: texts on the Video screen get their accents back.
  • Video: the Transition presentation says it renders online and locally.
  • Video: Ultra mode starts at 720p (no more 480p).
  • API and Connectors: these tabs present Ultra mode and 4K video, reserved for Pro license, that a server or AI agent can request. The API tab shows a ready-to-copy example of an Ultra 4K video.

v0.6.8

  • License: the Pro offer card details what it adds, with 4K video rendering (Ultra mode), cloud render API key, and MCP connector to create images and videos from Claude, ChatGPT or your AI agent.

v0.6.7

  • Video: "Next shot" really extends the clip. The next shot continues movement from the last second of the previous one, instead of starting again from a frozen image: no more stop at the cut. During a sequence, framing and size follow the previous shot and duration is fixed. Sound restarts fresh at each shot: keep the same ambiance in your text so the cut sounds as seamless as possible. On your card and online.
  • Video: online render modes are now called "Fast" and "Quality", without engine name. Nothing changes in rendering or prices.
  • Video: new "Ultra" mode, the finest rendering (water, foam, reflections, textures), online. The button appears as soon as the service opens it, with its price. Longer to render than Quality. In Ultra, 4K is offered with Pro license (upscaled: Ultra 1080p enlarged on output), up to 20 s, next shot included.

v0.6.6

  • Simple creation screen: the welcome sentence shows image and video icons from the switch, and the input example no longer talks about photos (the switch chooses).
  • Gallery on phone: thumbnails take up the full width, and the Size slider chooses one, two or three columns.
  • Phone: the Settings tab > Connectors appears, with Claude, ChatGPT and other agents (copy the address, paste it in the agent).

v0.6.5

  • Gallery on phone: an Images/Videos toggle at the top, to find your videos; the "Open" button, which could open nothing on Android, no longer appears there.
  • Simple creation screen: during rendering, a spinner next to the send button, which becomes a stop button to cancel; no more Cancel button or sentence under input.
  • Phone: the render queue bar no longer appears at the bottom of the screen.

v0.6.4

  • Whats New window: it presents Pro license (API and MCP, connectors, upgrade without losing anything) instead of the lines from 0.6.0.
  • Free Pro upgrade: the block no longer announces credits, it gives Pro without credits (only the paid upgrade completes credits to 500).

v0.6.3

  • License screen: offline, the fallback price is the real full rate (42 EUR) instead of the old strikethrough launch price. Online, nothing changes: the price comes from the site.
  • License screen: the dollar amount (English, Russian) follows the ECB rate from September.
  • Components: names and descriptions that stayed in French in the other seven languages (Krea 2 RedCraft, prompt editing, LTX 2.5 files, demo characters) are translated.
  • Settings > Connectors (Pro license): connect your agent (Claude Desktop, Claude Code, Cursor, VS Code, Windsurf) to Tendre.AI in one click; the agent writes prompts, launches cloud renders and follows results with your account credits. For claude.ai and ChatGPT, the address copies and pastes in their settings. Each connection is an API key named after the agent, visible in the API tab. Each line says whether the agent really connected (last activity) or is still waiting; it turns green on its own after pasting.
  • Switch your license to Pro: the "Unlock Pro" button opens the purchase for THIS license (same ID, characters and credits kept), and the app rereads its license on return and startup: Pro activates without reimporting files.
  • API and Connectors tabs with a Personal license: the price of upgrading to Pro (the difference, not a new Pro price), or "Free" for early licensees during launch week.
  • Settings > API: a tab to control cloud rendering from your own server or agent (Pro license). Address and ready-to-copy examples; named keys with permissions (images, videos), daily credit limit and expiration date, shown once only, five at most, revocable; and activity: each render made by the API, its cost, its state, consumption per key. Keys and activity stay hidden as long as the API is not open on the server side.
  • Settings > API without Pro license (trial or personal license): the "Feature reserved" lock, like other reserved modes, with a button to the Pro offer, instead of an empty tab.
  • The editing screen is called "Video editing" in English and "Videobewerking" in Dutch ("Montage" was a false friend in these two languages).
  • Purchase screen: two offers side by side, Personal (42 EUR) and Pro (99 EUR early bird until November 21, then 149), Pro with 500 credits offered, API key and render priority. Price and offer status come from the site: the Pro card only appears when the site declares it open; the button leads to the Pro purchase page.
  • Activation screen (PC): under "Import my license", a link "I have an activation link" to paste the link received by email, like on mobile.
  • First launch assistant: on an unsupported card, the "cloud mode" card no longer repeats the explanation already given under the graphics card, and no longer mentions engine size.
  • Generate: under a second character LoRA, help gives the trigger word to write in the prompt; it wrongly said this LoRA was tied to no character.
  • Generate and Video: the price in credits reads just next to the button, instead of being pushed to the other end of the row.
  • Video: canceling a clip then launching one right away no longer makes the progress bar yo-yo, and no longer accepts a second click on Generate while the new clip renders.
  • The full display mode is now called Studio (Starter and Studio), not to confuse it with Pro license.

v0.6.2

  • Cancel really stops. A canceled render did not always stop: in some cases the clip restarted on its own, without voice, and the card stayed at 100% after the click. Cancel now cuts the render AND everything it would have restarted; a batch of images stops at the current image instead of posting the next one.
  • Local training frees the card before leaving, and returns it to the app after. It started with the image engine still loaded in memory, which overflowed the 8 GB card; and once done, failed or canceled, the engine is relaunched so the first image after does not land on a shut-down engine.
  • Generate and Create a video have the same row of buttons, in the same order, and the progress bar moved below, across the full width: the timer, batch counter and receipt stay to its right instead of crushing it.
  • A chat clip is an animated photo: the requested action must be in the frame and in the gesture, not only in the text.
  • Several clips in a row, each with its own text. While a clip renders, Add to queue prepares another: it leaves right away in the render queue, which chains them on your card and shows them in order. Do not confuse with Chain a shot, under the preview, which takes the last image of a clip to make it the start of the next: there it is a single scene in several shots.
  • Your renders on your card go one after another, never one on top of another. An image launched during a clip no longer interrupts it: it waits its turn. Training takes the card to its end, and everything after waits. The bar at the bottom left says what renders and how many wait; a click opens the full queue, on your card and online, with cancellation entry by entry. This answers a lived error: launching an image, clip and training all at once overflowed the card's memory.

v0.6.1

  • Chat photos follow what's said more closely. A garment you just removed is no longer redrawn in the next photo, and a scene that is setting up is no longer rendered as if the action had already started.
  • On phone, you finally see who you are talking to: the character's thumbnail and first name at the top of the conversation, its thumbnail in front of each of its messages, and the menu to clear the conversation returns to the top right, where it was invisible.
  • The free trial counter says what you have left, plus what you had at the start: two numbers and two icons separated by the green dot, instead of words. The full detail stays on hover. This is the space returned to the character's first name on phone.
  • A trained character also acts in Retoucher. Choose it and its face returns in the photo you edit, on your card and online.

v0.6.0

  • When you click delete, it deletes. Confirmation windows ("Are you sure?") disappear from the entire app: gallery, characters, components, voices, LoRA library, training, video migration, conversations. Only one remains, the one that costs money: unlocking a paid character requires confirmation and shows the price. It wasn't displaying anymore, and worse, the purchase went through without waiting for your answer; that's fixed.
  • Every button that costs credits says how many, right next to it. Generate an image or batch online, retouch, launch a creation in Starter mode, equip a conversation with photos or videos, build a character preview: the price displays before you click, as it already did for video. Nothing displays when nothing is deducted (rendering on your card, free trial). In exchange, the line "Rendering: Cloud GPU (credits)" disappears from the top of the Generate screen: it said the same thing, too far from the button.
  • The credit counter now follows the conversation. It stayed frozen during an entire exchange, paid photos and clips included, and only updated when you left and came back to the window: you could think nothing was being deducted. It now updates after each photo, each clip, and each unlocked character.
  • One image engine: Krea 2. SDXL and InstantID leave the app, with their models, settings, and training. The LoRAs from the old engine stay visible and deletable in your components, but their card says they no longer serve to generate: for faithful faces, that's a Krea 2 LoRA, on your card or online.
  • Setup offers functions, not brands. You check what you want to do (images, videos, a chat) and the assistant installs what's needed, if your machine can handle it. It talks about your machine, not your graphics card. One video locally only, the 2.5, in 480p, 720p, or 1080p.
  • Train your characters on your own card. At the Training step, a button "on my card" next to cloud, and both estimated times announced before launching (roughly 5 hours on an 8 GB card, 55 minutes online); the estimate adjusts to actual speed from the first steps. The workshop and its base model install with the "Train a character" function when your machine allows (8 GB VRAM and 32 GB RAM), or from Components. Training saves its state every 250 steps: a break picks up where it left off, never from zero.
  • The character is born at the first gesture: from "Complete" or "Continue", it exists in My Characters with its state (photos 3 of 15, ready to train, training in progress), and clicking its card reopens the assistant where you left off.
  • Lea and Lucas no longer download to your card: they arrive with setup, at the same time as the Krea 2 pack. If you delete them from your disk, the app won't offer them to you again.
  • Pocah leaves the studio to stay in the chat of licensed characters, with her face. She no longer installs on your machine and no longer generates in Create an image.
  • The demo runs on the same engine as licensed characters, and its badge announces the 5 trial images truly offered.
  • Chat takes the full height of the screen on desktop, and the full screen on phone. The portrait keeps its proportions.
  • The language detector no longer mistakes a French sentence for Dutch, and the app's language has the final word. Stage directions follow the conversation language.
  • A video's prompt accounts for what the other person is doing: no more solo scene glued to a two-person scene.
  • Online studio images are no longer offered to licensed characters: each image costs 1 credit from the first of the day. Chat keeps its 4 free photos a day, and the demo its 5 trial images.
  • While building a character sheet, a real Cancel button, a real progress bar, and the count of remaining images next to it.
  • Quality tier and character strength are set where you use them, in Generate, instead of forcing a round trip through Settings. Generate and Retouch also announce in how many pixels they'll render, and High size says what it actually does. On free trial, High size wears its lock: rendering is capped at 1024, and a portrait requested in 2K came back square.
  • Chat casting stays in chat: in Create an image, only Lea and Lucas are offered. The duplicate that showed them twice in the list disappears.
  • On first launch of a new version, a summary of what it brings, with a link to the full list on the site. Once only, not on every opening.
  • Engine startup shows a progress bar instead of a frozen line, and the time left for a character sheet no longer counts model loading in each remaining image.
  • On phone, the license selector no longer grays out the file to choose.
  • The backup chat responds correctly when our main server is unavailable.

v0.5.12

  • Pocah joins the Far West cast, with her own trained face: she generates like the others, on your card as well as online.
  • You can rate chat photos. A heart under the ones you like, open to all. The thumbs down, however, is reserved: saying an image is off requires knowing what it should have looked like, and this feedback serves to correct course.
  • Second take: below a photo you reject, one line is enough to say what's wrong. The scene is redone taking your comment into account.
  • Video trial plays in 720p, on our best engine. In exchange, fast mode and 480 and 1080 resolutions become reserved for licensed users: their lock is now on the button in question, and clicking opens the offers. The Create a video tab, meanwhile, loses its: it announced a closed section when the video trial is offered.
  • In character training, images made by Complete really go with your photos. The screen counted them, training didn't receive them: online it refused to start for lack of enough images, and on your card it deleted them before training without them. The framings you import also come back when you reopen the card, instead of leaving three empty slots.
  • Photos you import all display, instead of falling into a broken thumbnail from the second one on. And if one is in a format the app doesn't know how to display (iPhone HEIC, TIFF, RAW), it tells you and offers you a JPEG, instead of staying silent on a photo that might not serve training.
  • The bar of an online render announces the current step instead of just a timer.
  • When you write your name or theirs in a scene, the image receives it in correct English: "he pins her", never again "him pins she". And a two-person scene is no longer marked as a solo portrait because the other was named by their first name.
  • The anal act is only announced, and its model loaded, when the scene writes it. A scene that says vaginal is no longer prefixed with the opposite.
  • Characters answer in your language, and stay there. On a long conversation played in a language other than the app's, a response could leave in the app's language: what the character remembers of you is now written in the conversation language, and if a response does leave in the wrong language, it's redone, then translated as a last resort.
  • Online, everything goes through Krea 2: SDXL is stopped there, character training included. An SDXL LoRA you already trained remains usable on your card, it's simply no longer accepted online.
  • You can place an image in your scene in Krea 2: a piece of clothing, an object, another person. Click it to pick it up, press Delete to remove it. This was until now reserved for SDXL.
  • When a retouch can only be done online while you're on your card, the app switches and tells you. It did so silently, and you found out the charge afterward.
  • If the Krea 2 pack is installed, it's chosen on opening. A perfectly equipped machine would leave for online and pay credits for nothing.
  • Create a character works again. Photo creation had no effect, and characteristic creation demanded you generate an image first, when you only wanted to describe a character.
  • In training images, Redo cancels and redoes: you go back to the previous framing instead of launching until you land on it again.
  • Canceling an online render responds right away, instead of leaving the button frozen for the time of a server round trip.

v0.5.11

  • Lea and Lucas are shipped with the Krea 2 pack: you already have two characters to generate from first launch, without training anything. Without the Krea 2 pack they don't download, their models would serve nothing. Lucas also works online.
  • The gallery opens on your disk: an Open button shows in the explorer the folder you're viewing (images or videos, subfolder included), the breadcrumb root does the same, and refresh becomes a small icon to the left of the title.
  • Lea and Lucas generate even without a license. The character selector was closed on trial, which amounted to hiding the app's argument from those coming to discover it.
  • In Create a video, character and voice are chosen side by side.
  • Starter respects the format you request in your sentence: writing "in portrait format" or "in 16/9" is enough, no need to touch the setting.
  • In character video training, keeping the original voice from your clips is presented as a CHOICE, not as an absence of voice. The summary says what will happen instead of asking for something that shouldn't happen.

v0.5.10

  • Your voices and characters finally display in Create a video when the engine is on Local. They only appeared after a round trip through Cloud, giving the impression an imported voice wasn't taken into account.
  • During the few seconds of local engine startup, the generate button says it's waiting for the engine instead of staying gray for no reason. The rest of the form is usable: you can write your shot while waiting.
  • Automatic duration arrives on your own card: like online, the engine chooses the shot length itself based on what you write, up to twenty seconds. The small file it uses to decide installs on its own at launch if you already have video locally. If your card doesn't keep up on a long shot, the app tells you and suggests shortening or going online.
  • A character whose LoRA you already own can be used online: before each cloud render, the app checks if it's available on our servers and, if not, PROPOSES to send it there announcing its size. Nothing is sent without your consent: the file belongs to you.
  • Without this send, the render stops with the steps instead of the lora_unknown error code, which left no way out.
  • Krea 2 becomes the default engine when its pack is installed, and the chosen engine no longer resets on its own.
  • When the switch is on Local but your machine can't do the render, the app no longer sends it online on its own: it asks you, announcing the batch price, and sends nothing if you refuse. The fallback stays available, it becomes a choice.
  • A character's activation word can be corrected by hand in their card. This word is the one their LoRA was trained with; it's written nowhere in the file, and the app was guessing it from the filename. When it got it wrong, the LoRA would load silently and the render came out a stranger. On import, the app first asks for the real word from your online space if the LoRA comes from us.
  • A character created by characteristics and linked to a LoRA keeps their age and gender before the activation word. Without them, the model filled the gap its own way.

v0.5.9

  • Character training finally displays its progress: the counter stayed at 0/1000 from start to finish for image training, on work running normally.
  • The announced time left is realistic. An image training of about thirty minutes could announce over four hours.
  • Training steps (prep, captions, masking, upload) display in the interface language instead of French.
  • Installing the voice pack no longer kills the progress bar of an ongoing install: it disappeared in the middle while the download went on.
  • Jane's portrait displays again in character choice.
  • Each generation displays the time passing during render, then the duration it took. Images and videos.
  • When a render leaves for online for lack of the Krea 2 pack, the app says NOW which files are missing, instead of announcing an absent pack with no further detail.
  • Landscape format in High (2K): the image came out in 4:3 instead of the wide format requested. Fixed server-side, already in effect without an update.

v0.5.8

  • Setup no longer blocks on PCs whose Windows can't open our archive: the app uses 7-Zip if present, and the archives themselves have been remade in a format every Windows knows how to read (nothing to do on your end, already online).
  • A failed install SAYS so, and says why: until now the progress bar disappeared with the error, and a click with no effect was all you saw.
  • Installing video: the folder choice opens in the right place, the one holding TendreEngine. Choosing the engine itself no longer installs a second engine next to it.
  • Components: five entries displayed in French regardless of language (photo analysis, ControlNet Canny, Krea 2, and Krea 2 Edit).

v0.5.7

  • Your characters train in VIDEO: the app builds their clips, learns their face and voice, and gives you back a model that animates them. The journey goes end-to-end on your own graphics card, except the final training, too heavy for a consumer machine, which goes online (300 credits per 1000 steps).
  • Character voice works locally: the chosen timbre is applied to the clip on your own graphics card, without going through the cloud. The voice pack (2.5 GB) installs on its own at launch when local video is installed, with its progress at the bottom of the menu.
  • Your video character is used everywhere: online as well as on your own card. Its file lands on your PC at the end of training, and a button retrieves it if you change machines or trained it before this version.
  • The word that calls your character is finally displayed on their card, and writes itself in the prompt when you choose them. A character has one word only, whatever engine their LoRA comes from.
  • Video training: the character's photos enter the learning alongside clips. A photo is sharp where a clip is blurry: it's what anchors the face.
  • The screen no longer stays silent while training runs: it displays it, says what's left, and finally lets you stop it, even after a PC restart.
  • Video training: dataset clips are visible, play on hover, and come out, like photos. Clips you import count toward the total and reduce what's left to build.
  • Video training: the dataset also builds on your own graphics card when the video engine is installed.
  • Video training: voice is always learned when the dataset allows. A catalog voice applies to the whole dataset; with no voice chosen, the real voice from your clips is learned, and no speaking clip is built.
  • Training: the screen truly follows the training in progress (phase, elapsed time, estimate) instead of staying frozen on its first line.
  • Training: creating a character starts from a blank slate, instead of keeping the previous one's photos.
  • Training: the voice catalog fits on one screen, its families are translated, and a single title heads the page.
  • Video: the preview plays the converted clip, not the raw engine render.
  • Components offers video engine 2.5 for download, instead of 2.3 which no longer installs anywhere.
  • A complete Krea 2 install no longer gets taken for an empty one: it wrongly switched renders to the cloud, and each image cost credits when the card could produce it.
  • Navigation finally speaks the app's language in eight languages: three entries (Image to text, Image to image, Roleplay) stayed in English.
  • Image training: a Male LoRA no longer produces women in dresses, Complete completes instead of redoing everything, an image already on disk is no longer declared missing, and the dataset plan gains a full step back.

v0.5.6

  • Retouch: what's unavailable is grayed out and explained, instead of hidden or renamed. Adding an element says where to find it when it's missing, and the element stays selectable after being placed.
  • You can place an image in a scene, and Krea 2 blends it in.
  • The video module announces the real price and takes up width correctly.
  • Training: the number of steps becomes a slider locked to save points, instead of a free field.
  • Jane joins the character cast.
  • The grayed SDXL button makes sense without explanation.

v0.5.5

  • Voice by character is back on the new generation video engine: the selector reappears, and the chosen timbre is properly applied to the clip.
  • The setup wizard announces the right video engine: it offered LTX 2.3 when it installs 2.5, and underestimated the download (30 GB announced for 33).
  • The offer to install video no longer comes back when it's already installed: it only recognized the old engine, and reappeared at every launch for those who just installed the new one.
  • The technical text that goes with companion photos no longer displays during setup.

v0.5.4

  • Engine update can no longer carry off your models. When a package is rejected by integrity check, the old engine takes back its place WITH its models: they're only transferred once the new package is valid. Previously, a rejected package left with already-transferred files and left an empty engine, no message.
  • A broken update no longer replays identically: the downloaded package is now identified by its source and deleted if it fails, instead of being reused indefinitely.
  • Setup announces the right video engine: the startup wizard, the install notification, and the Components tab say LTX 2.5 and its real size (~33 GB), instead of announcing the old version. The notification also recognizes a 2.5 install already in place instead of offering to re-download a pack already present.
  • Free video trial shows the same interface as the licensed version: engine choice (LTX 2.5, automatic duration) is visible, only Turbo stays reserved for licensed users, with explanation on hover.
  • The Camera Movement menu offers simple values on the LTX 2.5 engine: a list of classic movements built into the scene description. The catalog pilot stays on renders that take it into account.
  • A character chosen locally no longer stays silently selected when switching to cloud, and the technical reminder of the local engine leaves the Video screen.
  • Transition also works locally with the LTX 2.5 engine: the tab respects the Local / Cloud GPU switch like the rest of video, instead of forcing cloud.
  • On free trial, Transition and Montage really open: those two tabs showed an empty screen. Transition is bound by the trial quota, Montage is local and free.

v0.5.3

  • The local video engine goes to LTX 2.5: sharper image, better character fidelity, still with sound. New installs get it right away; existing installs have an update button in the Video tab, which keeps the old engine on standby during download (~33 GB) and only removes it after a first successful video. A "tight disk" path frees the old engine first, your choice.
  • Characters trained in cloud are also used locally, with nothing redone.
  • The video character arrives in video creation in LOCAL: if one of your characters has a video LoRA, a "Character" selector appears and their identity is applied to the clip, with a strength slider.

v0.5.2

  • Important phone fix: the update window no longer overflowed the screen when notes were long, and the install button became unreachable. Notes now scroll inside, the button stays visible.
  • Chat photos are all framed by the director, and when they can't, the reason is stated instead of silent fallback.
  • Chat: speech registers apply to all characters.

v0.5.1

  • Android app no longer closes on startup.
  • On phone, no sound plays when the device is in vibrate or silent mode: alerts follow the phone's settings, like any app, and it's the system notification that alerts you.
  • The character selection screen adapts to your phone: your active adventures appear first, search collapses into a magnifying glass, and your identity is set from the pill in the top left, world by world.
  • The Create button becomes an arrow placed in the text area, on phone and desktop alike.
  • The button to return to the demo is no longer stuck under the phone's status bar.
  • The free trial focuses on three characters: Lea, Lucas, and Kara Vega. Other worlds remain visible and unlock on activation, rather than opening halfway.
  • Images generated online all go through Krea 2, including retouching and character training. The SDXL engine remains available on your computer, with your LoRA, as before.
  • Generating a video now writes only one file: the clip. No images are created alongside it.
  • Duration automatically resets to automatic as soon as the online engine is active.
  • Resume restores all a clip's settings, but not its starting image: you need to upload it again.
  • Fix: after a long absence, the next photo could show your character in a way that nothing in the conversation had prompted.

v0.5.0

  • New default video engine: sharper image, long shots in one take, sound included, and a duration the model can decide based on your text.
  • Duration is set by dragging the value with your mouse, or left on automatic.
  • The price of a video no longer depends on its duration: only format and sound matter. A long clip now costs the same as a short one.
  • Transitions from one image to another also move to the new engine.
  • Resume, from the gallery, restores everything that produced the clip: settings, character, style, and even the starting image.
  • Simplified mode benefits from the new engine and lets the model choose the shot duration.
  • Character selection now fits in your hand: on small screens, worlds become a menu, search expands, and cards align in pairs.
  • Chat photos are now framed by an outside perspective that reads the entire scene before describing it: the location follows the conversation, your own gesture counts, and the image shows the moment that just happened.
  • Desire is read in the scene, not in the words: a curious question or a kiss on the cheek no longer shift a conversation, and a scene that heats up doesn't wait its turn.
  • Chat videos last ten seconds with sound, for everyone, including free trial.
  • Fix: the activation page announced ten times too many videos offered.
  • Fix: a calm scene could veer into adult territory because the character sheet mentioned certain physical features.
  • Fix: technical copies of images no longer accumulate next to your videos.

v0.4.0

  • New world: the FAR WEST, and its first resident, Silenne. A mining town from the 1870s, a saloon, a staircase.
  • New character selection screen: all worlds on a single page, filtered by world, search by name, and each character becomes a portrait.
  • Your identity per world: say who you are in a world when you start a conversation there, and find the sheet on the right to change it whenever you want.
  • Streamlined creation screen: the advice below the field disappears, the Create button becomes clean and moves to the right.
  • Characters truly die when the scene calls for it: a gunshot at close range is no longer dodged.
  • Chat: no more confusion between you and the character in two-person scene photos.

v0.3.4

  • Internal preparation fixes, no visible changes.

v0.3.3

  • Streamlined creation screen: the advice below the field was disappearing behind the two photo and video buttons that already said the same thing. The Create button becomes clean and moves to the right, at the end of the gesture.
  • First Android version to receive the new character selection screen.

v0.3.2

  • New character selection screen: worlds are stacked on a single page, filterable by chips, and each character becomes a portrait (profession in a pill, name as overlay, difficulty in bars).
  • Your identity per world: say who you are in a world when you start a conversation there, and find the sheet on the right to change it whenever you want.
  • Silenne, first resident of the Far West, is announced: her sheet arrives soon.
  • Loridia becomes Farmer, Lyra becomes Hunter.
  • Chat: no more confusion between you and the character in two-person scene photos (a female body is no longer described in masculine form, and vice versa).
  • Skin markings (abs, chest tattoo) no longer display when the scene is viewed from behind, which was distorting anatomy.
  • Training: character type now accepts 48 characters instead of 24.

v0.3.1

  • YOUR PAID CHARACTERS FOLLOW YOU. A character unlocked on your computer now appears on your phone too, and comes back after reinstalling the app (previously purchases were only remembered on the machine where you bought them). Past purchases are retrieved automatically.
  • NEW CHARACTERS CAN ARRIVE WITHOUT AN UPDATE. The character roster is now served by our servers: a published character shows up on computer AND phone without waiting for a new version. With no internet, your roster stays intact.
  • PHONE: the scenes screen didn't scroll before. Worlds spilled outside a fixed frame, the first was cut off at the top and the last was unreachable. Tiles now take the full width and the list scrolls.
  • PHONE: the credits banner no longer covers the title of screens and doesn't overflow on the left. The Local / Cloud GPU choice disappears from it: there's no graphics card to drive on a phone.
  • Sasha is no longer announced by a category on her card: she says it herself, in her own words, when you ask her.

v0.3.0

  • LORIDIA, a medieval peasant, opens a new universe. She carries her identity model, her world in its cover, and her story turns: at the heart of the village festival, her family vanishes in a silent white flash and leaves only a golden feather. The moment is marked by the scene, not by a message counter, and the message where everything changes is marked with a golden border in the conversation.
  • MAJOR FIX: what you write between asterisks finally goes to the image engine. Your action stayed in French and in the first person in the prompt, or it drew nothing, so the image showed the scene without you. It is now translated, put in the third person, agreed, and your description takes the place of the pronoun: you who act in the image.
  • Your chat videos last ten seconds instead of five, with sound, and intense scenes go through the dedicated anatomy model, which was announced but never reached quick rendering.
  • The video price is what is charged. The button announced 40 credits when the server took 4: there is now only one number, served by the server, and it is 20.
  • Photos finally follow the story when the scene changes location. The character went downstairs, settled by the fire, and three photos in a row still showed them in the attic.
  • The character is allowed to have people around them: their family, the table, the villagers. A duplicate lock erased them all.
  • MEGAN finds their face again: their identity model was published but never authorized on the server side, and their photos changed faces for five days.
  • SASHA is no longer announced by a category on their thumbnail: they say it themselves, in their own words, when asked.
  • Roleplay: your name and the character's name no longer go to the image engine, where they invited a stranger into the frame. A response no longer starts with a title ("# Response"). Returns ("change character", "universe") are now real buttons.

v0.2.9

  • MAJOR FIX: the character's photo finally shows what they just wrote. It was described BEFORE their response, so it always showed the previous moment: they were coming out of the shower in a t-shirt and the photo showed them still in swimwear, they were starting the car and the photo showed them in their kitchen.
  • Roleplay: one framing per photo only. Two contradictory instructions (wide shot and close-up) produced impossible bodies; likewise, a body described from behind with a face facing forward. An action is now framed wide enough to see it, instead of a close-up on the face that hid it.
  • Roleplay: a refusal written by the model in their photo description is no longer drawn.
  • Roleplay: the scene described by the character is no longer automatically replaced in certain actions, or completed with words they did not write.
  • Training: Krea 2 by default, and the base model no longer gets lost when switching tabs (training would run on SDXL without saying so). Character type is now free: woman, man, or anything else you enter.
  • Training: startup shortened by around twenty minutes on the server side, models no longer need to be downloaded each time.

v0.2.8

  • NEW CHARACTER, LEA: the library assistant in the cast now has her own trained identity model. Her face stays consistent from photo to photo, and her profile has been expanded: twenty-four years old, literature degree since June, first apartment above a bakery.
  • Roleplay: your character can DIE. An act that would kill a human really kills them, they play their death, and the story stops there. One outcome only, restart a new story. Detection only triggers on a direct act: a threat, an argument or an image is not enough.
  • FIX: a conversation in English switched to French mid-scene and never came back. The language safeguard was listening to the app's language instead of the one you write in.
  • FIX: a trained character finally uses THEIR own identity model. The app preferred facial recognition whenever there was a photo, and displayed the strength of a system that was not being used.
  • Roleplay: the character no longer refuses to play a dark scene by breaking character, and can no longer rewrite an act to undo it. Narration entirely in English in the middle of a French response is now caught too.
  • Roleplay: a trans character keeps their role. The reminder that set their anatomy presupposed a position, and imposed it over and over.
  • Two new image styles, Smartphone Photo and Direct Flash: the flaws of a real photo taken in a hurry, where all other styles lean toward polished. Reserved for scenes without a trained character.

v0.2.7

  • Fix: character could appear doubled in photos. The sentence locking their identity at the head of the prompt started with an article, and an image engine reads an indefinite article as the announcement of a second person to draw.
  • Roleplay: a conversation line can no longer be mistaken for a scene. "I can't picture it" became a location, translated then pasted into all following photos.
  • Roleplay: the relationship no longer restarts from closeness after a fight. The level reached is retained, and reconciliation brings you back to it: a quarrel is a detour, it doesn't erase what was built.

v0.2.6

  • Roleplay: the character's name no longer leaks into the prompt for their photos. "Lina B3YND" was including its surname, and an unknown word stuck to the subject was degrading the render.
  • Roleplay: Lina B3YND is now the cover of the Cyberpunk universe.

v0.2.5

  • DEMO OPENS UP: 5 images and 1 video offered EVERY DAY (instead of 3 images once and for all), and the core app becomes accessible without a license. My Characters, Image from a Character, Transition and Montage are open; only character training, scenes, LoRA library and voices remain restricted.
  • ROLEPLAY (formerly Chat): the tab opens on a choice of UNIVERSE (Real, Cyberpunk, Fantasy) and free casting opens COMPLETELY, Kara Vega included. No character is reserved for licenses anymore: the limit is the daily quota, not a closed door.
  • NEW CHARACTER, LINA B3YND (Cyberpunk universe): architect by day, painter in private, equipped with white ceramic prosthetics edged with gold that she never hides. Sapiosexual: you win her over with the rightness of an idea, not compliments. Free like the rest of casting.
  • Roleplay: casting displays instantly. It was waiting for complete character sync, which downloaded up to 700 MB on first launch; portraits are now embedded as thumbnails.
  • Roleplay: two-person scenes are no longer restricted (the image could refuse to display the other character), illustrated mode no longer loses photos, and a translation refusal can no longer end up drawn in the image.
  • Megan now has her stable face across all her photos, like Kara.
  • Fix: character sheet title displayed "Personality and chat" instead of "Character sheet".

v0.2.4

  • Fix Chat: "one photo per response" and "one video per response" modes work again (they went silent for licensed users after chat photos moved to cloud). In video mode, each response produces a clip, even if the character describes a photo.
  • Official characters (Kara, Lea, Megan, Lucas, etc.): character sheet in READ-ONLY mode with "Official" badge on their cards. You can use them everywhere (images, chat) but can't rename, edit, export, or delete them.
  • Fix: character sheet title displayed "Personality and Chat" instead of "Character Sheet".

v0.2.3

  • NEW CHARACTER: KARA VEGA. 24-year-old netrunner, dominant, sarcastic, and self-made. First character powered by a Krea 2 LoRA: maximum face fidelity across all her photos, guaranteed signature silhouette. Free, in chat.
  • CHAT UNIVERSE: the Chat tab now opens on your world choice (Real, Fantasy, Cyberpunk), with large animated tiles. Each character lives in their universe; upcoming ones will already be sorted.
  • GALLERY: click an image to select it and activate selection mode (Shift for a range). Videos gain the same full mode: batch deletion and organization.
  • Generate: "Character Strength" slider for LoRA characters (Krea 2, and cloud), from 0.30 to 1.20.
  • CHARACTER SHARING: the .tendrechar file now carries the Krea 2 LoRA and complete personality (import sorts the LoRA to the right location automatically).
  • Chat: LoRA character photos can no longer lose their face (their appearance description no longer seeps into images), long memory writes in your language, and display cleans up (no more technical notes in brackets in bubbles; scene directions remain).
  • Sheet: Save stays on the sheet, Back warns if edits aren't saved, and clicking the portrait imports a new profile image.

v0.2.2

  • ONE LORA PER ENGINE ON EACH CHARACTER: a single character now carries its SDXL LoRA, its Krea 2 LoRA, and its video LoRA, each used by the right engine. The character and LoRA lists in Generate show only what is compatible with the chosen engine, and an imported LoRA sorts itself into the correct slot.
  • KREA 2 IN CLOUD, COMPLETE: your Krea 2 character works in Cloud GPU too (its LoRA loads on our servers), and scenes are rendered by a dedicated model instead of being redirected to SDXL. What works locally works in the cloud.
  • Character sheet: voice joins the Identity card with a direct import button (the imported file also feeds the voice library), an "Import a LoRA" button places the file in the right spot on the character, and the base model card disappears.
  • Generate: a line under the title reminds you when rendering goes to Cloud GPU. The negative prompt is no longer sent to the Krea 2 engine, which does not read it: shorter prompts and clearer logs.
  • CHAT: photos of a character with Krea 2 LoRA are rendered by Krea 2, with much higher face fidelity; conversation intensity chooses the model (reserved or intense).
  • Chat: display fixes (a photo prompt could show raw in the bubble) and scene fixes (a word of enthusiasm like "good idea" could be mistaken for a location and stick to all photos).

v0.2.1

  • Character Training: when you enlarge a generated image from the dataset, its prompt displays and you can EDIT it before regenerating. Fix the text of a failed pose instead of replaying the same prompt hoping for a better seed; follow-up attempts keep your text.
  • Fix: in Character Training, the identifier was written incorrectly as soon as you edited it manually (the suffix kept adding on each keystroke, impossible to delete). It now cleans up normally.

v0.2.0

  • CHARACTER TRAINING ON KREA 2: the training form now offers the Krea 2 base in addition to SDXL. Training runs on our servers (48 to 80 GB GPU), about 35 minutes, and the resulting LoRA is used in the Generate tab with the Krea 2 engine. Tested recommendation: LoRA strength around 0.80, and prioritize variety of photos over quantity.
  • LOCAL / CLOUD GPU SWITCH: a toggle next to your credit balance chooses where everything generates. Local = your graphics card, free; Cloud GPU = our servers, faster even with a good card, your PC stays available. It controls images, retouching, and video (the Video tab loses its own selector).
  • CHARACTER SHEET redesigned: two columns, portrait with hover regeneration, status badge, trigger copyable in one click, linked Identity / Model / LoRA / Images / Training cards, and Cancel-Save bar. Deletion moves to the bottom of the page, and the folder icon on a LoRA opens its file in your file explorer.
  • RETOUCHING: when your card lacks memory for a large image, a window offers to switch to the cloud, showing the price, instead of a simple error message. Nothing goes to our servers without your consent.
  • License: price displays in dollars when the app language is not European.
  • Generate: the negative prompt field grays out with the Krea 2 engine, which doesn't read it; describe what you want directly in the prompt.
  • Trial version: new screenshots in the premium features screen (LoRA library, scenes, voices).

v0.1.71

  • Automatic character preview: at the end of training, the character thumbnail generates automatically with its LoRA, and through the correct engine (a Krea 2 LoRA was being routed through SDXL and rendering an unknown character). If preview is impossible (video LoRA), a dataset image serves as the thumbnail.
  • Revamped character profile: photo at the top, associated LoRAs (image and video), training dataset characteristics now stored and displayed, base model chosen from a list. Preview and Training buttons leave the cards for the profile, keeping cards clean.
  • Deleting a character truly deletes its card: its LoRA file stays in the Library, without reappearing in "Recreate character". Unlinked LoRA cards now have an X to remove them the same way.
  • Free video trial usable everywhere: the Cloud engine in the Video tab is no longer locked without a license, the free video generates directly (quota shown under engine choice).
  • Trial version: the restricted features screen now shows screenshots and visuals of what the license unlocks, large and clickable.
  • Chat: the character replies in the language of your last message, and follows your language changes, instead of drifting to English.
  • License: price, promotion and its end date, and offered credits display live from the website (200 credits offered, roughly a hundred videos).
  • Translations: analysis of training photos displays in your language.

v0.1.70

  • LORA ON THE GENERATE SCREEN: the "Generate LoRA" tab disappears, its settings join Generate. An "Add a LoRA" link under the character, as many LoRAs as you want, each with its strength, and it works for both SDXL and Krea 2. The LoRA ADDS to the character instead of replacing it: identity on one side, style on the other.
  • LoRA: the character's trigger word is added automatically to your prompt. Without it, a LoRA is loaded but stays silent, and failure looks like failed training. For an imported LoRA that belongs to no character, the screen reminds you to write it.
  • Character training: the supplement board finally contains close-ups and expressions, a smile and a laugh included. It only produced full-body shots, where a face takes up too little space to be learned.
  • Training: the next step opens as soon as its content exists, without waiting for the Continue button. And restarting an image now sorts it into its board's folder.
  • FREE TRIAL: three images offered EVERY DAY instead of a one-time three-image trial, non-cumulative. The remaining count displays next to the reload button.
  • Chat: the character pauses before starting to write, the time a person would take to read your message. The response came instantly, which gave away the machine.
  • Video: the Cinematic category moves to the top of the styles list.
  • Interface: single color for everything related to credits, and brighter on colored buttons.

v0.1.69

  • VIDEO FROM TEXT ALONE, LOCALLY: the Video tab no longer requires a starting image when generation runs on your card. Describe the scene, LTX 2.3 films it, with sound. The restriction came from the previous engine, which couldn't do it.
  • Video: the style list is now limited to the 21 styles actually supported by LTX 2.3 (Animation, Stylized, Cinematic). Old custom styles, which had no effect on this engine, have been removed.
  • Character: new Height characteristic, translated into descriptive vocabulary and scale index for wide shots. Age is capped at 18 years minimum.
  • Generate: when your card can generate locally but no model is installed, a banner tells you and offers installation. Before, images went to the cloud and consumed your trial without explanation.
  • Installation: the Starter / Pro choice disappears when your card can generate locally, you go straight to Pro. It only remains for machines that have cloud only.
  • Library: character LoRAs now have a type, Image (SDXL) or Video (LTX 2.3). A video LoRA can no longer be chosen by mistake as an image character.
  • Fix: your personal LoRAs reappear in their Library tab, they were displaying under Krea 2 since 0.1.68.

v0.1.68

  • MODERNIZED ALL-CLOUD MODE: Krea 2 becomes the default photo engine on machines without compatible GPU (cutting-edge photorealism, native 2K in High mode, same price as before). Adult scenes and LoRA characters continue to automatically use SDXL, no setup needed.
  • NEW GENERATION ATTACHMENT: the attached photo is now EDITED according to your instruction (Krea 2) instead of being redrawn. The person and scene are preserved, only your request is applied. The intensity slider disappears, the adjustment is automatic.
  • Gallery cards always show the engine actually used for each image.
  • Components: face detector labels now display in your language (they remained in French).

v0.1.67

  • INSTRUCTION-BASED RETOUCHING (Krea 2): new mode in the Retoucher tab, next to the brush. Load an image, describe the edit in English ("add a knight riding the dragon"): the entire image is reworked while preserving the scene, and depth order (front / back) is finally understood. Local if the Krea 2 pack is installed, otherwise cloud rendering at the cost of one photo (licenses).
  • The Krea 2 pack installation now includes editing (~20.4 GB total), and already-installed engines receive the editing module automatically at next startup (45 KB).
  • Retoucher: screen reorganized. The mode choice (SDXL Masking or Instruction) moves to the top, the model choice goes down into the Masking mode box (it only acts there).
  • Fix: "Improve face" was failing (error 400) on installations without the Character pack. The face detector is now downloaded automatically on first use (~52 MB).
  • Fix: generating with a character created from a LoRA was failing without the Character pack. The LoRA now carries the identity locally (as it does in cloud), and a character per photo without the pack displays a clear message instead of a technical error.

v0.1.66

  • KREA 2 TURBO, the next-generation image engine, arrives in Tendre.AI: cutting-edge photorealism, native 2K in High mode. New "Image Engine" choice at the top of the Generate tab. Local if the pack is installed (approximately 18.6 GB, offered at installation), otherwise transparent cloud rendering at the cost of one photo.
  • Installation: Krea 2 at the top of the models grid. Juggernaut exits the assistant (still available in the Components tab).
  • Retoucher: multi-generation (1, 2, 4, or 8 images) now works in retouching too: multiple variants of the same area, all in the gallery, the last one applied to the image.
  • Retoucher: Removal mode can no longer invent content when you don't ask for it (neutral fill and reinforced safeguards).
  • Transition: choice of Local or Cloud directly on screen (morphing also runs on the local model), all resolutions become available again, and Transition is never sent to the Turbo worker.
  • Video and Transition in cloud: minimum duration lowered to 3 seconds.
  • Image cards now indicate the engine actually used (no more SDXL metadata on a Krea 2 render).
  • Refreshed visuals: new home image, new Megan card, softened labels.

v0.1.65

  • UNIFIED ENGINE: photos and video now run on the same engine (TendreEngine v2). Existing installations upgrade automatically in the background, models preserved, with automatic rollback if anything goes wrong; the old engine is kept as a safety backup.
  • NEW: character consistency per photo (InstantID) is now truly available on your machine. Optional at install, or one click from the photo creation page (about 4.7 GB).
  • The unified engine opens the door to next-generation models (coming in future releases).
  • Older graphics cards (GTX 900/1000): the app automatically uses the cloud, no unnecessary local setup.
  • The setup assistant displays the character portrait in the waiting conversation.

v0.1.64

  • Video: the Turbo model is now offered by default (and first in the selector): we're proposing the fastest and cheapest option first.
  • Video: resuming a clip from the gallery in 720p or longer than 10 seconds automatically switches back to LTX 2.3 Full (Turbo would have reduced its resolution and duration).

v0.1.63

  • NEW PRICING (already active on our servers): - Cloud photo: 1 credit (0.05 euro) instead of 2, in the studio and in conversations. Offered quotas don't change (5 free conversation photos per day with a license). - NEW: Turbo video. A lightweight model that renders in tens of seconds, at 480p, up to 10 seconds, WITH sound: 2 credits for a 5-second video (0.10 euro), 4 credits for 10 seconds. The cheapest way to get a video with sound on the market. - Quality video (full resolution, up to 20 seconds, voice, face restoration) keeps its pricing.
  • Video: at the top of the Generate tab, choose your model: LTX 2.3 Full (the complete model: full resolution, up to 20s, voice and HD) or LTX 2.3 Turbo (fast and economical). Turbo replaces the old "Draft".

v0.1.62

  • The file name shown by Windows during installation becomes TENDRE. Nothing else changes in this version.

v0.1.61

  • SIMPLIFIED CHARACTERS: single tab instead of four. The "Create a character" button opens the path choice (LoRA training, reference photo, characteristics), each card has its Training button that goes straight to the form, and creation by training happens in place (name + type, character is created on launch).
  • Chat: gone are the nearly identical photos that repeat. Each photo varies the pose or angle; objects no longer accumulate from one photo to the next.
  • Chat: the character no longer comments on its own rules in its responses, and language instructions have been consolidated across all 9 languages.
  • Retoucher: object removal becomes a toolbar tool (wand): masking the object is enough, the formula is composed for you.
  • Local video: the Sound checkbox is back. A silent clip also generates faster (the audio track is no longer calculated at all).
  • Training: progress displays discreetly at the bottom of the sidebar, you can keep working during cloud training.
  • Creation by characteristics: 8 more hair colors added (platinum, auburn, white, silver, pink, blue, purple, green).

v0.1.60

  • Training: single model selector in the same place for both Local and Cloud (under engine choice). Locally it lists your installed models, in Cloud it shows the 4 base catalog models.
  • Retoucher: the work area is bounded by a frame, the image is centered and fits within the window height (vertical images used to float and overflow below the screen).
  • Retoucher: scroll wheel zoom anchors exactly under your cursor (the target point used to drift, especially on narrow images).
  • Retoucher: Undo and Redo moved to the top of the toolbar.

v0.1.59

  • Cloud Training: single model choice on screen. The selector was sharing display with the Generate tab's choice, causing two model menus to contradict each other.
  • Chat tab: first response arrives faster. The conversation engine warms up when you open the tab, just like video did already; your first message used to wait up to a minute for startup.
  • Retoucher: tools now stack vertically along the preview (they were eating into image height), and zoom buttons are tighter.

v0.1.58

  • NEW: train the character of your choice on our servers (40 credits). A toggle between This PC / Cloud in the Training tab. In cloud: about 25 minutes, and your PC stays completely free during that time, no blocking screen, you keep generating. You can even close the app, tracking resumes when you reopen it. The LoRA installs itself afterward. Quality verified identical to local training, and even a notch above: it works at 1024 where 8 GB cards max out at 768.
  • Cloud training: choice of base model among the 4 in the catalog (Versatile, Neutral Photo, Cinematic, Artistic), even if the base isn't installed on your PC. A message will alert you that you'll need to install it to generate locally with it.
  • Training becomes accessible to licensees without an NVIDIA card: the trained character is used in companion photos and in Generate, via the cloud.
  • At generation, the training base of the character's LoRA is used automatically: no more mixing bases that damage likeness. Intentional exception: at the companion's hardcore tier, the dedicated model keeps control.
  • Cloud video: 1080p without audio goes up to 20 seconds, as does audio (it was capped at 5 s, an inconsistency).
  • A LoRA's file name is no longer inserted into the prompt in place of its trigger word.
  • Cloud generation: when our image servers are saturated, your request waits instead of failing.
  • The app now knows all graphics cards on the machine, not just NVIDIA.

v0.1.57

  • Chat: the character no longer comments on its own instructions in its responses.
  • Chat: emojis described in text are now displayed correctly instead of written out.
  • Chat: character profiles are written in English, like the rest of their instructions. This was the last block of French remaining, and it was pulling responses toward French in other languages.
  • Chat: Candy talks about her banana without holding one in every photo. It was a conversational habit that had become a permanent prop.
  • The prompt assistant is available to everyone, with nothing to install.
  • The app monitors your graphics card and suggests the right engine when it changes. Switching cards no longer leaves an incomplete installation without telling you.
  • Components: a menu per category instead of one long list, and an imported model appears in its group and nowhere else.

v0.1.56

  • Language fix, the real one. The character was responding in French regardless of the chosen language, and the previous version wasn't enough: its internal instructions were themselves written in French, hundreds of words, against a single line asking for Italian or Spanish. The model was following the dominant language of what it was reading. These instructions are now in English, a language it doesn't confuse with any conversation language.
  • Chat: conversation and its photos go entirely through our servers. Nothing left to download to chat (the 8 GB engine is no longer required), no more generation failures due to a missing model on disk, and your graphics card stays free for your images.
  • Chat: Sasha is listed as a trans woman in her profile, which is what her character has been from the start.
  • Fix: image generations were failing due to an error introduced last night on the server side. Fixed overnight.

v0.1.55

  • Fix: the character was responding in French regardless of the chosen language. Two distinct causes, both fixed: its welcome message ignored the translation, and its language instruction only forbade English, never French in which the app talks to it.
  • Chat: 5 free photos per day with a license, instead of none. Text remains free and unlimited.
  • Models you import now finally appear in selectors and in the components list. They were installed correctly but the app only showed its own catalog, so nothing seemed to be happening.
  • Model import: the type is recognized in the file (image model, LoRA, VAE, ControlNet) and you confirm it before installation, so it goes in the right folder.
  • Components: the Refresh and Import buttons are combined at the top of the page.
  • Chat: image instructions tightened. They had accumulated to the point of diluting each other, which weakened the ones that mattered.

v0.1.54

  • PRIVACY: your text translation no longer goes through Google. Until now, to make the models understand a request written in French, it was translated by an external service from your machine. That's over: translation is done by the engine installed on you when it runs, otherwise by our own servers. No third-party service, and Google's address is removed from the app's network permissions.
  • Chat: in French and English, no more translation needed, the conversation is understood directly. The seven other languages too, without the slightest network call between your message and the response.
  • Chat: the character no longer switches to English when automatic images are active. It was writing its image description in English and continuing on its momentum.
  • Settings: reorganized into sections with a menu, instead of a long scrolling page. The image and video output folders are now at the top, where you look for them.
  • A Discord button in the sidebar, to join the community.

v0.1.53

  • NEW: Upscaling tab. Doubles the resolution of an already made image without regenerating it, so it doesn't change. Processes multiple images in a row, and works on an image imported from outside too.
  • NEW: chat with Lea or Lucas during setup. The engine download is long, might as well use the time to get acquainted. Text only, no license needed.
  • Retouching: narrow areas are finally sharp. A mouth or eye in a wide shot was processed at the scale of the entire image, which gave a soft result no matter the setting. The masked area is now retouched at full resolution then restitched, seamlessly.
  • Retouching: element integration uses settings suited to an OBJECT (more flexible stroke, free colors). The old preset aimed at a logo and stuck organic objects without explanation. For a real brand, recheck "Keep exact colors".
  • Gallery: multiple selection with Ctrl and Shift, move to an existing or new folder, and batch deletion.
  • Gallery: media displays in rows of identical height with a size slider. Comparing two generations becomes much more readable.
  • Resume: the visual style is recovered with the rest. It used to reset to neutral, so the render wasn't reproduced faithfully.
  • Training: character type accepts creature, animal, object and a free field. A non-human character was tied to human, which hurt its resemblance.
  • Video: the AI assistant also reworks the audio line, instead of erasing it.
  • Video: a start image too heavy no longer fails rendering (it is adjusted to clip resolution).
  • Chat: in trial version, the character now expresses what holds them back and offers the license at the right moment, instead of a technical message.
  • Generate: the two header bars contradicted each other, they are removed. Functions reserved for GPU explain their absence on hover.

v0.1.52

  • NEW: Transition accepts sound. The Sound checkbox was hidden in this tab, it's now available like in Create a video.
  • NEW: camera movement in Transition. You can impose a pan, orbit or aerial descent during the passage from one image to another, instead of following the trajectory chosen by the engine. Great for going from an aerial shot to a close-up without breaking the style or the character.
  • Gallery: multiple selection. A Select button turns thumbnails into checkboxes, with Ctrl to add an image and Shift to take a whole range. Ctrl or Shift also works directly on a thumbnail, without going through the button.
  • Gallery: organize your images. Once selected, you move them to an existing folder or to a new folder you create on the spot. Ideal for preparing a set of images before training.
  • Gallery: media displays in horizontal rows of identical height, instead of offset columns. Comparing two generations becomes much more readable, and no image is cropped.
  • Resume: the visual style is now found along with the rest. Until now it went back to neutral, so the render was not reproduced faithfully.
  • Training: character type is no longer limited to woman, man or person. Creature, animal and object are offered, and a free field lets you write what you want (orc, dragon, robot...). A non-human character was previously tied to human, which damaged its likeness.

v0.1.51

  • NEW: cloud video camera movements. Choose left or right orbit, forward or backward tracking, left or right pan, aerial descent, crane lift, or follow shot, and the camera executes the actual movement instead of interpreting it from text. The menu includes 14 movements, the 9 core ones and 5 others described in the prompt.
  • Camera movements work at all durations and with sound.
  • Duration as a slider from 4 to 20 seconds in 1-second steps, replacing fixed tiers.
  • Camera movement quality greatly improved: references no longer let foreign elements into your clips (no watermark overlay, no actor from the original shot), and both pans were redone.
  • Transition: new Transition Force slider, from soft fade to exact arrival on the end image. Lower it when your two images are far apart and the passage feels jarring.
  • Transition: the end of the clip is no longer frozen on the arrival image, and the duration obtained now matches what you requested.
  • Transition: the Continue imported video link is now available there, to chain one shot after the previous one.
  • Transition limited to 15 seconds: beyond that, the render came out black.
  • Image: Landscape and Portrait formats now go to 16:9 and 9:16 to match video. A generated image can now be animated without losing an edge to reframing. The app warns you if the end image of a transition doesn't have the right ratio.
  • Chat: on a PC without a compatible card, chat automatically goes through the cloud (no more proposal to download the 8 GB local engine). An already installed local engine stays in use.
  • End of free trial: clear message with license purchase button, instead of a technical error code.
  • Security: local engines now accept only requests from the application (a web page open in your browser can no longer control them).

v0.1.50

  • NEW: local video WITH SOUND (LTX 2.3). Generate clips of 5 to 20 seconds with sound, directly on your card, free and unlimited. Replaces the old silent video engine.
  • Cloud video: up to 20 seconds with sound, and 1080p for short clips.
  • NEW: TendreEngine, the unified engine in a single clean folder. Installation 5 to 10 times faster (dedicated download server + multi-connection downloads, speed displayed).
  • Video installation runs in the background: start creating while it downloads (discrete bar, cancelable, resumes on its own after an interruption).
  • Create (licenses): attach an image by paperclip or drag-and-drop to transform it into an image (img2img) or animate it into video.
  • If local video is not installed, the app offers it to you on startup (one click).
  • Redesigned wizard (feature lists, local or cloud video choice), clearer warnings (GPU memory, starting image too small), window maximized on startup.
  • First-launch terms of use are translated into 9 languages.
  • Pro mode remains available even without a local model: generation then goes through the cloud (no more forced switch to Starter if you chose Pro).
  • Support rule clarified: local generation on NVIDIA RTX series 20 or newer (8 GB and up). Older cards (GTX 9xx/10xx) automatically go through the cloud, where generation is much faster for them.

v0.1.49

  • NEW: Pro mode without GPU. If no capable NVIDIA card is detected, Generate and Retoucher automatically go through the cloud (image and brush retouch), with a clear banner. Element integration remains reserved for local GPU.
  • Chat: character no longer invents your name. Automatic extraction is hardened (a false name memorized could come back up in conversation).

v0.1.48

  • Components: import your own models in one click. An "Import a model (.safetensors)" button copies the file (e.g. downloaded from Civitai) to the right place, no need to navigate folders anymore.

v0.1.47

  • Chat: photos sent by your character now cost 2 credits. The cloud image in the Generate tab remains free.

v0.1.46

  • Cloud image is FREE for now (0 credit): take advantage to generate without GPU.
  • Retoucher (inpainting and element integration) is now accessible WITHOUT a license: retouch your images locally, freely.
  • Retoucher is translated into 9 languages of the application.

v0.1.45

  • NEW: Retoucher becomes a real little editor. A toolbar above the image (Move, Brush, Mask, Eraser), undo/redo action by action (Ctrl+Z / Ctrl+Alt+Z), and a bigger image on screen.
  • NEW: integrate an element (logo, badge, object as PNG) into an image. Place it directly on the image: drag to move, a corner to resize, Ctrl + corner for perspective, the circle above for rotation. The AI then blends it into the material (texture, folds, light) while keeping its exact shape and colors.
  • Mask tool cuts out the element to make it pass behind an object or person (e.g. a shoulder in the foreground). Option: automatically confine the element to the person below it, with mask preview.
  • Paint a zone and generate without a prompt: the zone is redone in agreement with what surrounds it. No prompt (positive or negative) is added hidden in Retoucher: what you type is exactly what is sent.
  • The chosen style also applies to retouches: the zone is rendered as photo, drawing or painting depending on the style.
  • The prompt assistant (AI button) now clearly shows why it does not answer (engine not installed, busy...) instead of failing silently.
  • Character training becomes "LoRA Training", featured just under My Characters.
  • LoRA generation: the "delete" button has been removed to avoid accidentally erasing a LoRA.

v0.1.44

  • NEW: give a VOICE to your characters. Import an audio sample into the Library (Voice tab), and your spoken videos adopt this timbre, mouth synced. In cloud video, a voice selector; in chat, each character keeps its voice.
  • Cloud video much more responsive: the engine starts much faster (models built into the server).
  • Chat: character no longer teleports itself into its favorite scene. The location of the scene stays the one of your exchange until you change it.
  • LoRA Library: a file's type can be changed freely, and the LoRA add menu is sorted by type.

v0.1.43

  • Chat: character expression truly follows the scene. No more permanent smile when the scene calls for a serious look, a pout or closed eyes; default smile stays on neutral scenes.
  • LoRA Library: a file's type can be changed freely (choosing "Scene" no longer fixes the type permanently), errors are displayed, and known files are recognized regardless of name case.
  • LoRA generation: the "Add a LoRA" menu is sorted by type (character, scene, style, action), like the main selector.

v0.1.42

  • New LoRA Library (sidebar, Library section): all your LoRA files with their type, trigger word and links; import, reclassify, delete.
  • LoRA generation: stack multiple LoRA (scene, style, action...) each with its own strength.
  • LoRA are sorted by type (character, scene, style, action) in all menus.
  • Characters: deletion by trash icon on each card (including LoRA characters).
  • Chat: if a required LoRA is missing on your PC, the photo automatically goes through the cloud.

v0.1.41

  • Mobile chat: videos finally save (default folder automatic, like photos).

v0.1.40

  • Chat: no more bare character photo when conversation is not at that stage (character description is cleaned before the image, in addition to the existing safeguard).

v0.1.39

  • Chat: notification on device when a character replies while the app is not on screen (adjustable in Settings).
  • Chat: the reply arrives and notification goes even if you quit the app while the character is writing.
  • Chat: invitation to unlock the app displays as a normal message in the conversation.
  • Mobile: updates are offered automatically on app startup.
  • Generation: cloud photos and chat replies much faster (renewed engines).

v0.1.38

  • Chat: you can scroll up to read the conversation while the character is typing (scrolling no longer forces you back to the bottom).
  • NEW: a "Check for updates" button in Settings to install the latest version whenever you want.
  • Activation screen redesigned: the license offer and benefits are clearer.
  • Improved license activation reliability.

v0.1.37

  • Chat: your very first message no longer fails at startup. While the engine wakes up, the app waits and reconnects automatically, never showing an error.
  • Chat: the "typing" indicator is now animated while waiting for a response.
  • Video: the engine selector clearly displays "Local" and "Cloud".
  • New application icon.

v0.1.35

  • Chat: two photo and video buttons to the left of the input to automatically receive an image (or clip) with each response, whenever you want.
  • Chat: a notification sound when your character replies (with a button to turn it off).
  • Chat: videos arrive faster (the video engine prepares itself as soon as the conversation warms up).
  • Candy: her photos now look like her in cloud mode too, and her appearance has been refined.
  • Chat: smoother flow when you chain photos and videos, and successive photos really vary when the scene changes (no more nearly identical images).
  • Chat: "show me", "I'd like to see you" and their variants trigger photo sending correctly in all languages.
  • Chat: several display fixes (video loading that could disappear, inconsistent scene) and consistency between text and image.
  • Your creations are stored in Documents\Tendre.AI by default on new installations.

v0.1.34

  • NEW premium character: Candy, 26 years old, nuclear and solar safety engineer, playful. First companion with her own identity model (LoRA) downloaded automatically when unlocked: her photos really look like her.
  • Chat: more HUMAN characters: each has their own pace, makes you wait with charm if you go too fast, and reacts with wit if treated like an object, never breaking the game.
  • Chat: the cost of a clip is displayed on the "Turn into film" button; your name is pronounced only in the first video; at warm levels the character truly executes the action from the image, with confidence.
  • Your creations (images, videos, trainings) are now stored in Documents\Tendre.AI by default (new installations): no more unfindable hidden folder.
  • Cloud chat: new more reliable engine (no more empty responses, faster startup).

v0.1.33

  • Chat: characters develop depth and react with personality. Videos have real continuous action and spoken dialogue adapted to the moment (your name at warm levels, scene intensity at intense levels).
  • Sasha: 100% anatomy consistent with their trans identity, in their words and in their images.
  • Chat: change your name by simply asking the character ("call me..."), in all languages.
  • Chat: no more responses mixed with English and messages that disappeared for no reason (refined locks).
  • Interface: in Starter mode, settings and help move to the bottom of the sidebar, like in Pro mode.
  • Cloud: the chat engine now turns on automatically on demand and turns off after 30 minutes of inactivity (first response of a session in 1 to 3 minutes, instant after).

v0.1.32

  • Chat: the character types at human speed (natural typing with variations) instead of displaying their response all at once; their photo prepares while they type and arrives right after, like a real exchange.
  • Chat: no more responses written "in your place" (the character no longer replies by playing your role) and no more overly long responses.
  • Chat: text smileys become emojis (<3, :), ;p...) both ways, and characters can slip them into their messages, like in real texts.
  • Characters: premium character profiles now follow app updates. Lyra now fully lives in her high fantasy world (no more anachronisms), platinum blonde wavy hair, slender figure.

v0.1.31

  • Chat in Starter mode: conversations run IN THE CLOUD, no download or graphics card needed. Same characters, same natural feel: works instantly, even on a basic phone.
  • Chat in demo (no license): chat free with Lea and Lucas, 20 messages per day. Asking for a photo, video, or another character shows a prompt to activate a license right in the conversation (one-click purchase).
  • The Chat tab is always visible. In Pro mode without a chat engine, a button installs it in one click from the tab (~8 GB, one time only).
  • Character selection: available characters display first, locked ones show a clear lock icon and premium characters show their price.
  • Reliability and security of the free trial strengthened on the server side.

v0.1.30

  • Chat: at higher intensity levels, images automatically switch to the advanced model for superior rendering.
  • Chat in DEMO (no license): chat freely with Lea and Lucas (unlimited text, local engine). Photos, videos, and other characters lead to license activation.
  • Chat: shorter and faster replies (reinforced prompt + adjusted cap), no more stray stage directions like "(Photo of me...)" in chat bubbles, and the "Photo with each reply" checkbox reappears automatically when the local engine finishes starting.
  • Chat: "Clear conversation" also erases the current memory summary (no memory can resurface afterward), and the Chat tab displays on launch (it sometimes waited for a tab switch).
  • Cloud: evening capacity increased (2 warm workers, up to 4 at peak) for instant renders even during high traffic.

v0.1.29

  • Chat: two new PREMIUM characters unlockable with credits, Sasha (100 credits) and Lyra (200 credits). An unlocked character stays unlocked forever (even after reinstall).
  • Chat: ILLUSTRATED mode (local engine required): check "Photo with each response" and the character accompanies each of their messages with a photo of their scene, generated free on your graphics card.
  • Chat: photos flow with real continuity (same location, same lighting, the scene evolves as the conversation goes), the requested framing is respected, and duplicate faces are corrected.
  • Chat: change your character's attitude by asking them ("be more compliant", "be dominant", "be shy"...): it persists until you say otherwise. "Clear conversation" restores their original character.
  • Chat: the character knows your local time and date (coherent greetings and activities), no longer invents shared memories, describes their photos faithfully, and no longer overreacts to ordinary questions.
  • Chat: click a photo to open it fullscreen (scroll to zoom), side portrait reduced, conversation and thread images enlarged. Megan is now a nurse, swimmer, and more playful.
  • Unlicensed: local VIDEO generation is now free and unlimited (like images): only the cloud engine requires a license. Your GPU, your creations.
  • Pro mode: if your NVIDIA card is compatible but the local engine isn't installed, a banner offers one-click installation.

v0.1.28

  • Security: strengthened content protection filter. It now applies uniformly to ALL generations (images, videos, LoRA, retouching, chat), local and cloud, with a clear message when a prompt is blocked.
  • Cloud reliability: automatic failover to a second datacenter if the main one is saturated (rendering continues instead of failing at peak times).

v0.1.27

  • Chat companion: your characters can now send you VIDEOS of themselves (roughly 5 seconds, with sound) when you ask ("film yourself..."), plus photos.
  • Chat: "Turn into film" button on each received photo to animate that photo into a short sound clip.
  • Chat: the character gets to know you. They ask your name as you chat, remember it (for all characters), and address you personally.
  • Chat: in Pro mode with local engine, character photos are generated on your NVIDIA card with face locking (higher fidelity), free; the cloud is used otherwise.
  • Characters: Megan's figure refined (athletic, swimmer).

v0.1.26

  • Chat companion: your characters can now send you PHOTOS of themselves directly in the conversation ("send me a photo", "I want to see you", "show me..."), generated in the cloud while they write their reply. The requested framing (full-length, selfie, etc.) is respected.
  • Chat companion: much more natural conversations. Short and calibrated responses to your messages, no more role-play breaks, the character keeps their identity, gender, and your language even in long conversations.
  • Chat: the chat cast is now made up of the app's official characters (with their backstory), separated from characters you create for image generation.
  • Chat: you can interrupt the character's response by sending your next message, and freely scroll back through the conversation while the character is typing.
  • Create: each image or video in the stream can be deleted (cross in the corner), from disk and gallery.
  • Credits: displayed amounts move to the new scale (10 credits = 1 video of 5 s, 2 credits = 1 photo); existing balances have been converted automatically, value unchanged.
  • Cloud reliability: a render lost in the queue is automatically relaunched, with no action from you.

v0.1.25

  • Starter mode simplified: on first launch, choosing Starter goes straight into cloud mode, with no engine installation or model downloads. Perfect to get started in seconds.
  • Clearer choice screen: Starter mode (100% cloud, no installation) and Pro mode (local generation, NVIDIA GPU over 8 GB required) are now explicitly described.
  • Auto-switch to cloud if no image model is installed: if the local engine is present but has no checkpoint (nothing to generate locally), the app automatically switches to cloud mode instead of failing. Once a model is added, it goes back to local.

v0.1.24

  • App digitally signed (verified publisher: PifletteProd). Gone is the Windows "unknown publisher" warning on install: Windows now recognizes Tendre.AI as trusted software.

v0.1.23

  • Starter mode: you choose Image or Video with two icons left of the prompt (no more auto-detection, you decide).
  • Starter mode: videos are much better quality. An image is generated first from your text, then animated, which matches what you ask for much better.
  • Machines without compatible NVIDIA GPU: lightweight install, you go straight into cloud mode without downloading the engine (7 GB saved).
  • Scrollbar adapted to dark theme.

v0.1.22

  • Starter mode: videos are now generated in 720p vertical with sound (ideal format for mobile and social media), and faster.
  • Starter mode: your creations stream is kept when you switch tabs and come back.

v0.1.21

  • Fix: download errors (error 520) on large renders, especially HD video: images and videos are now retrieved in binary (lighter and more reliable) instead of text format that would saturate on large files.

v0.1.20

  • Activation screen: the "Import my license" button is now a real, clearly visible button (blue outline).

v0.1.19

  • Starter mode: your prompt is automatically translated to English before generation. You can write in your language, and models (trained in English) then give much better results.
  • "Activate my license" button more visible in the free trial end message.

v0.1.18

  • Free trial: without a license, 3 images and 1 video are free in the cloud to discover the app (ideal for machines without a compatible NVIDIA GPU). After that, a license unlocks unlimited generation locally or via cloud credits.
  • Starter mode display fixes (progress bar, assistant layout).

v0.1.17

  • New Starter mode: a single field to describe what you want, the app generates the image (or video) and figures out the rest. Choose on first launch, changeable anytime in settings.
  • Image generation in the cloud: machines without a compatible NVIDIA GPU can now generate via the cloud.
  • Character duo: two coherent characters in the same image.
  • New Sound tab: automatically adds a soundtrack to a silent video.
  • Reorganized gallery: folders shown in grid, with subfolders for videos and editing.
  • Clearer install assistant: models shown as thumbnails with sample image.

v0.1.16

  • Support for OLDER graphics cards (GTX 900 / GTX 1000, TITAN X, Quadro M/P, MX): the install assistant detects these cards and automatically installs the engine version that supports them. Before, generation would fail on these machines ("no kernel image").

v0.1.15

  • IMPORTANT fix: when generation failed on the engine side (missing model, memory, etc.), the app displayed NOTHING (no image or error). The exact reason for failure now displays clearly.
  • Translations completed in 9 languages: style selectors (image and video), groups and chat model in the Models tab, base status in the sidebar.

v0.1.14

  • The app now requests administrator rights on launch (Windows prompt): ComfyUI installation works in all folders, including Program Files.
  • Setup assistant: if the chosen folder is protected or inaccessible, a clear message explains it and suggests a better location (instead of a technical error).
  • Chat: the welcome message from provided characters displays in your app language (and changes with it).
  • Chat: the personality of characters provided with the app is locked (yours stay editable).
  • Activation screen: direct button to buy a license on tendre.ai.
  • Gallery: Resume and Delete buttons redesigned (clearer, accidental deletion protected by its discreteness).

v0.1.13

  • IMPORTANT fix: on first launch (or after updating from an old version), the database could fail to open ("duplicate column") and block the app. Migration is now protected.
  • Fix: if the saved ComfyUI folder no longer exists (different machine, disconnected drive, cleanup), the app reopens the setup assistant instead of failing with "access denied".
  • New: "Pool house" scene FREE, installed automatically (downloaded once in background). Your character in a real place from day one.
  • New: BATCH generation. Choose 1/2/4/8 images per click, generated in sequence with variations, saved on the fly.
  • New: QUEUE generation. While one image renders, write the next prompt and add it to the queue: everything chains automatically.
  • New: RESUME button in gallery (images and videos): reload in one click the prompt, seed, character, scene, format and settings from a creation to replay it or vary it.
  • Anonymous counter of active users (random identifier, no personal data).

v0.1.12

  • New: IMAGE and VIDEO STYLES. A selector of moods (cinematic, portrait, etc.) that enriches your prompt in one click, on Generate and Video sides.
  • New: TRANSITION (morphing). Give a starting image and an ending image, the app creates the video that interpolates from one to the other. New dedicated tab.
  • New: VIDEO CHAINING and SEQUENCES. Chain your shots (the last image of one clip becomes the start of the next) and auto-arrange them in a named folder (01_name, 02_name...). You can also import a video to continue it.
  • Gallery reorganized: nested folders (parents and children) with breadcrumb, instead of listing everything.
  • Video: 8 s duration now available for 480p with sound.
  • Video: the app remembers your engine (cloud or local) and fixes the "not ready" banner that stayed displayed wrongly in cloud mode.
  • Settings: license import button (file .tendrelic).
  • Various interface fixes (checkboxes, alignments, browse button).

v0.1.11

  • New: SCENE creation (places). Describe a scene (interior/exterior, moment, mood) or import photos, and the app trains a LoRA that keeps that place coherent across all your generations. Thumbnails in the scenes list.
  • New: PUT YOUR CHARACTER IN YOUR SCENE. Scene selector at generation (InstantID character or LoRA character), with a strength slider, to blend identity and place.
  • Interface: forms in two columns (Generate, Video, LoRA) to use screen space better, settings accessible via a gear wheel at the bottom of the sidebar.
  • Removed custom video LoRA field and audio mode description, mouse shortcut legend under the retouch area.

v0.1.10

  • Cloud video credit pricing adjusted: 720p now costs double 480p beyond 5 s (10 s and 15 s), reflecting real compute cost. The 5 s clip stays at 1 credit regardless of size.

v0.1.9

  • Cloud video: 720p now renders in true high definition even on long clips (10 s and 15 s), instead of being downscaled to 480p.
  • Cloud video: removed custom LoRA field (security: no longer reference free files on remote volume).

v0.1.8

  • Chat companion: chat with your characters via 100 percent local, unrestricted text AI.
  • 5 base characters shipped with the app (3 free, 2 to unlock), face and personality included.
  • Inpainting: retouch a specific area of an image from a simple prompt.
  • Image preview: zoom with scroll wheel and pan with middle click.
  • Integrated prompt optimization assistant at generation.
  • Video gallery: prompt copy button.
  • Reorganized menu (Characters first, collapsible sections).
  • Interface fully translated in 9 languages.

v0.1.7

  • Previous reference version (auto-updater, Ed25519 license, LTX cloud video, LoRA training).
Windows · Mac · Cloud or local

Generate privately. Tonight.

Download tendre.AI for free and generate without limits, privately, on your own GPU.

Download for WindowsDownload for Android (.apk)Download for Mac
OS Windows 10/11GPU NVIDIA 8GB+VRAM 12GB recommendedDisk 20GB

Mac: Apple silicon (M1 or later), macOS 13 or later. Intel Macs are not supported.

Local image generation on Mac needs an M4 with 24 GB of memory, or an M2 or M3 Pro, Max or Ultra with 24 GB. Below that, everything runs online with credits. Retouching and video run online on Mac for now.

Everything about tendre.AI for Mac

No NVIDIA GPU? Start 100% in the cloud: 5 free images and 1 free video every day. An NVIDIA card (8GB+) unlocks unlimited local generation.

Join Discord