Features · Compositions
Put it all together.
Compositions are where generated media becomes a finished piece: a video on a multi-track timeline, or an audio piece with its own speakers. Build it by hand, or describe it and let the AI chat draft it.
Video
Edit whole videos on a timeline.
A video composition is built for long-form work: titles, images, video clips, and audio arranged on a multi-track timeline and rendered to one finished file. The clips themselves come from the video workspace, your own footage, or a layer generated in place.
A multi-track timeline
Titles, images, video, and audio each get their own track. Drag to move, pull an edge to resize, drag up or down to change what draws on top, and snap cleanly to neighbouring clips, the playhead, or a clip's own true length when you're trimming it. Fades, volume, motion, and exact position live in the layer panel, or just drag and scale a title directly on the preview, with guides that snap it to the centre and the edges. A new composition starts empty (three lanes for titles, visuals, and audio at 1920×1080), and an empty image, video, or audio layer can be generated straight into, at the layer's own length.
AI drafts the whole cut
Tell the chat what the video is for and it plans the beats and lays them onto the timeline as real clips (titles, voice-over, music, sound effects), then generates the image and audio layers in parallel, so timings and title cards are playable immediately while the rest backfill. It picks one visual theme and holds every image to it, so the shots look like one shoot. Refine with follow-ups: nudge a timing, reword a title, quieten the music. Each request touches only what you asked about and undoes in one step.
Characters that look the same in every shot
The Cast tab holds the people, animals, and creatures that recur in a video. Give each one a look: a picture you upload, or one generated from a short description as a single portrait or as a sheet of several angles and outfits. Every image or clip that shows them is then generated with that look as its reference, so they stay recognisable from shot to shot, and each character can have a voice of their own for the lines they say. Ask the chat for a story and it casts the characters itself, making their looks before any shot that needs them.
Turn a still into a shot
Every layer has a properties panel: timing and fades, exact position and size, motion, volume and looping, and for titles the fonts actually installed on your machine. Image layers get one more: Make video. Describe the motion and your chosen model animates the still from that exact frame; the clip swaps in place, keeping its timing and framing, and one ⌘Z puts the image back. Video clips get the same two Extend actions as the video workspace, with the new footage landing right after the source on its track, as one undo step.
Render on your machine
Export at native size or 480p to 2160p, at up to 60 fps, as MP4, WebM, or MOV. The render runs locally with ffmpeg (your footage never uploads), progress is shown with a cancel option, and the finished file lands in your project folder like any other clip.
Audio
Write it, cast it, hear it.
An audio composition is built for podcasts, interviews, and narrated stories: a script read by several speakers, with music and sound effects mixed underneath.
A script that drives the timeline
Every line of the script is a clip on its speaker's track, so the script and the timeline always show the same conversation. Type a line, choose who says it, and generate it when it reads right. When a line's audio arrives it takes its real length and everything after it moves to fit, so nothing overlaps.
A voice for every speaker
The Cast tab gives each speaker their own voice model and voice, with a preview so you can hear it before you generate a line. One piece can mix voices from different providers, and renaming a speaker renames their track too.
Mix it like a multi-track
Speakers, music, and sound effects each get their own track that you can mute with one click, and every clip shows its waveform. Add a music bed or a sound effect from your files, or generate one into an empty layer, then export the mix as MP3, M4A, or WAV.
Chat
Describe it, then refine it.
Every composition has its own chat. Describe the video or the episode you have in mind and it drafts the whole thing: clips on the timeline, or a script with a cast. Then ask for changes in plain language. Each request touches only what you asked about and lands as one undo step, and the chat runs on your Chat model, local or cloud.
Workspace
The rest of the editor.
Undo, empty layers, your own files, aspect ratios, and exports: how the editor behaves around the timeline.
One undo historyEvery edit you make by hand and every change the chat makes lands in the same history, so ⌘Z always takes back the last step, whoever made it.
Generate into empty layersAdd an empty image, video, music, or sound effect layer and generate straight into it, at the size and length the layer already has.
Drop in your own filesDrag images, video, or audio from your computer onto the canvas or the timeline, and the file is copied into your project folder.
Change the shape laterSwitch a video between nine aspect ratios, from square to portrait, and every layer is rescaled to fit instead of stretched.
Plain files on your diskEach composition is saved as a file in your project folder, next to the images, clips, and audio it uses.
Exports stay on your machineVideos and audio mixes are both built locally, so the media in them is never uploaded anywhere to be rendered.
FAQ
Composition questions, answered.
How compositions work in CSuite: video and audio pieces, scripts and voices, exports, the AI chat, and where your files live.
A composition is where you put media together into a finished piece. There are two types: a video on a multi-track timeline, and an audio piece with a cast of speakers. You pick the type when you create it, and each type opens in its own editor.
No. The type is fixed when you create the composition, because each type is laid out differently. Your generated images, clips, and audio stay in your project folder, so you can reuse them in a composition of another type.
Each line in the script is a voice clip on its speaker's track, so the script and the timeline always show the same thing. Type a line, choose who says it, and generate it. When a line's audio arrives it takes its real length and the clips after it move to fit, so the conversation never overlaps. If you change a line after it was generated, it is marked as changed until you generate it again.
Yes. The Cast tab gives each speaker their own voice model and voice, with the same settings and voice preview as the audio workspace, so one piece can mix voices from different providers. A speaker with no model chosen uses the voice model picked on the Models screen.
Yes. In a video composition's Cast tab, give the character a look: upload a picture, or generate one from a description as a portrait or a sheet of several angles and outfits. Every image or clip that shows them is generated with that look as a reference picture. If the image model you picked can't take a reference picture, CSuite uses one of your models that can and says which; a video layer uses a video model that takes reference pictures, or animates a still made with the look. With no model that can use a reference picture, the shot is made without it and the chat tells you.
Only when you generate it, one line at a time or with Generate all, which voices every written line that has no audio yet or has changed. Typing and rearranging lines costs nothing. Cloud voices are billed by the provider at that model's rates, directly to your account.
A video exports as MP4, WebM, or MOV at up to 4K and 60 fps. An audio piece exports its mix as MP3, M4A, or WAV.
Yes. Every composition has its own chat. Describe the video or the episode you want and it drafts the whole thing: clips on the timeline, or a script and a cast. Then ask for changes in plain language. Each request becomes one undo step, and the chat runs on your Chat model, local or cloud.
In your project folder, as plain files on your own disk, next to the media they use. There is no cloud library and no sync.