Changelog
New features, improvements, and fixes. Shipped constantly.
We’re focused on making every video look and feel higher quality, with sharper visuals, smoother motion, and more polished output on the first try.
We’re testing a new, more affordable Miltos model that stays just as smart, so great results cost you less.
Miltos 6 is a new top tier for complex, story-driven videos: richer motion design, stronger storytelling, and more careful edits. Launch pricing takes 30% off its AI tokens.
Photos and logos you attach through the API or MCP, including Google Drive links, are now saved into your project the moment your request arrives. They are always there when you export, and a link we can’t open is flagged right away so you can fix it before anything is built.
Ask for captions on any video with a voiceover and you get word-by-word animated captions, whether the scene is fully animated or a silent clip you narrated over.
Save a character once and it shows up as the same person in everything you make, not just from scene to scene. Describe one, upload artwork you already have, or turn a photo into an avatar, drawn in one of our video styles.
The mobile editor now shows which project you’re in, and sharing or exporting it takes one tap instead of a trip through a menu.
Picking a narration language no longer narrows the voice list. Every voice in the library is available for every language we support, and the one you choose is the one your video is narrated with.
Long videos with lots of narration and sound effects used to bog Safari down while the timeline loaded. The editor now opens and stays responsive on Mac and iPhone, however many audio clips your project carries.
Hit play on your phone and the preview runs start to finish, music and narration included, so you can check an edit from your iPhone or iPad instead of waiting until you are back at a desk.
When your video walks through a product, the cursor now travels to the exact button or field it is using, and the click or typing happens right where it is pointing.
Ask in chat how to save, download or export your video and an Export button appears right there in the conversation, one click from your finished file.
Ask for a specific song and we now find the studio audio rather than a music video, so your narration isn’t competing with dialogue, footsteps and sound effects baked into the track.
Animate a logo, product shot or any picture in chat and the video now automatically takes on the shape of your image, so a tall logo stays whole instead of losing its top and bottom to a widescreen crop.
Ask Claude for a video and a live card appears in the conversation. It fills in with your scenes as they are made, then plays the finished video where you are already sitting, so you never have to ask "is it done yet" or go hunting for a link.
Spotted the wrong quality or frame rate the moment you hit export? Stop the render in its first 10% and your video export goes straight back to your balance, so you can fix the setting and start again without it costing you anything.
The editor now keeps your video the biggest thing on screen, however many tracks a project has. Tracks scroll in place instead of squeezing the preview into a corner, clips open at a size you can actually see and grab on a phone, and you can fold the timeline away entirely when you just want to watch.
Upload your logo as an SVG and the AI now reads the real shapes inside the file. Ask for the lines to fly in and assemble, or for the outline to draw itself on, and it animates your actual mark rather than an approximation of it.
Long, video-heavy projects now play in the mobile editor instead of sitting on a loading spinner. The preview loads whatever you’re watching first and pulls the rest in quietly behind it, so playback starts sooner, keeps going, and uses far less of your mobile data along the way.
Sound effects are used more sparingly and placed with more care, so your videos sound cleaner and the moments that matter land harder.
Starting a video is down to what matters, on the home page and in the editor alike: describe your video, pick a style, and go. Format, length, scene count and sound are worked out from what you wrote, and if something genuinely can not be inferred the AI asks you one quick question instead of guessing. Everything you used to set up front is still one tap away in the + menu.
If your images end up in documents, reports or handouts, you no longer have to resize and crop them onto the page. Pick A4 in the aspect ratio menu and the image comes out the exact shape of the page, ready to drop straight in.
That intro you nailed last month, the outro you use everywhere, the one scene that finally landed: move them between projects instead of rebuilding them. Drag a scene straight in from another project, copy and paste it between tabs, or send it to a project from the right-click menu. Its clips and its narration come with it, timed to where it lands.
Once every scene is in place, the AI lays the entire film out on one contact sheet and gives it an art director’s pass: empty or unfinished scenes, broken layouts, the same composition used twice, a jarring cut, a scene that drifts off your style. What it finds, it fixes before it tells you it is done, and you see the sheet it judged, right in the chat.
The chat now shows its working. Frames of your footage as the AI checks them, rendered previews of a scene it has just written, the transcript of what it heard, and the sources behind any web search — all as it goes, so you can follow the reasoning instead of taking its word for it.
Short clips, like a quick line of narration or a brief cut, now select and drag just like longer ones, including shift-clicking them into a multi-selection. Narrow clips used to respond only to their trim edges.
Ask for a video about something happening right now and MotionVid searches the web first, then writes the script from what it finds.
Every turn in the editor now carries its own credit counter, climbing while the AI works and settling on the final number when it lands. You can watch what an edit is costing while it runs, and stop it if it is going further than you wanted.
Developers now have a proper front door at motionvid.ai/api. Pick your client and get its exact setup, then connect Cursor or VS Code with a single click and start generating videos from your agent or your own code.
A new full screen feed at /discover, one video per screen, starting with the latest projects people have published. Every clip shows the prompt behind it, so you can see exactly how it was made and pick up its style in a click.
Attaching references is no longer one image at a time. Pick as many files as you like in a single trip to the upload dialog, or tick several images you have already uploaded and add the whole set at once.
Accented and non-Latin letters now render in the font you picked, everywhere. Turkish, Polish, Czech, Romanian, Hungarian, Vietnamese, Greek and Cyrillic text stays in one consistent typeface across the editor and the finished video.
Finished videos are no longer final. Send a plain-language instruction ("swap the logo", "shorten the intro", "fix the map in scene 2") and the same AI editor behind the app applies it to your existing video, keeping everything that was already right. Works from the REST API and from AI assistants like Claude connected through MCP.
Maps with city pins now match your video exactly, whether it is landscape, portrait, or square. Coastlines keep their true proportions and every pin lands precisely on its city, in any format you publish.
Zoom and pan effects now play back fluidly in the editor preview. Effects ramp in and out without interrupting playback, so what you scrub is what you ship.
AI Video now casts your story before it films it. You see a portrait of every character and approve the look before a single clip is made, and those same faces carry through the whole video. Films also run up to two minutes now, with every shot directed like real camera work.
The composer now opens with four clear ways to start: AI Video, Motion Graphics, Single Graphic, and Edit My Video. Each mode shows its own animation styles and sets everything up for you, so you get the right kind of video without touching a single setting.
Photo-real, lifelike characters now animate more dependably on Animora Pro. We automatically route these portraits to the best engine behind the scenes, so more of your image-to-video generations finish on the first try.
Connect your own provider accounts and route eligible generations through them. Add and manage keys straight from your account menu.
Build MotionVid into your own apps. A new API lets your tools create and track videos, and AI assistants like Claude can now make videos for you directly.
Add text that appears exactly when you want it in a video, and start new projects from a simpler, more focused screen.
Uploaded videos now start playing almost instantly, and the cancellation flow offers real help and credits instead of a dead end.
When editing an uploaded video, the AI sometimes stopped after placing just two or three graphics. It now adds overlays throughout, wherever they belong, like stats, quotes, intros, and chapters.
Upload a video and layer AI-generated motion graphics, overlays, and cuts directly over your real footage, all built right into the editor.
Higher plans now come with unlimited exports, upscales, and background removals, plus bonus credits when you pay yearly, so you can create as much as you want without watching your balance.
Drop in audio-only assets, control how videos fit their frame, and use the app in Swedish.
The AI soundtrack could stop a few seconds before the last scene. It now always stretches to cover the full length of your video.
Every project can score itself with a custom AI soundtrack, talking-head clips keep their voice, and videos featuring real people are handled with extra care.
A refreshed landing showcase for Animora 2.0, and you can now drag videos and images straight into the timeline, even an empty one.
Reuse a whole set of reference images across remixes, and usage meters now show what you have left rather than what you’ve used.
Manage your subscription, pricing, and cancellation in one place, and remixes now better understand what you want to change.
A new engine behind our Animora videos for higher-quality generation, with 4K upscaling for your finished clips.
The editor now opens instantly, and you can upload your own images and videos into a media library to drop straight into projects.
Create animated maps with pins and connecting lines you can edit, great for showing locations, routes, and coverage areas.
Swap a scene’s image, drag a scene edge to change its speed instead of trimming, and remove backgrounds right from scene settings.
We improved how each video is generated behind the scenes, so your results come out looking better with less trial and error.
Everything above is live. Describe a video and it builds itself.
