Video editor
On the Video tab of the work screen you can edit a short video yourself or hand it to your assigned AI. It works well for vertical videos like Shorts.
What can you do?
- Upload videos, photos and audio files, place them as clips, then reorder or trim them.
- For each clip, choose the speed, whether to keep its own sound, and how it moves to the next clip (cut or crossfade).
- Add subtitles and music, or type a script and create narration with matching subtitles.
- Create a new video source from a scene description with Generate video.
- When it’s ready, use Export finished video to get a video file (mp4).
While the assigned AI is editing you can only watch, and you can edit again when it finishes. To stop it midway, press Stop AI editing.
Photos work as clips too
Upload a photo (png, jpg or webp, up to 30 MB) and it is placed as a 3-second clip. Select the clip to set the following in the settings panel on the right.
| Setting | What it does |
|---|---|
| Fit: Show all | Shows the whole photo or video without cropping, leaving black where there is space left |
| Fit: Fill frame | Fills the whole frame, and anything that spills over is cropped |
| Crop position | With Fill frame, a slider picks which part to show |
| Slow zoom | Photo clips only. When on, the photo slowly grows larger |
Photos have no sound, so the “clip’s own sound” setting doesn’t appear for them. The assigned AI can make the same settings.
Check before you export
The Export finished video button shows how many things are worth checking. Open it to see problems like these ahead of time.
- Narration or subtitles that run past the end of the video and get cut off
- Two narrations, or two subtitles in the same spot, that overlap
- Subtitles that show too briefly or touch the edge of the screen
These are only heads-ups, so you can still export as is. Select an item to jump to that spot.
Costs and storage
- Creating narration and exporting a finished video use the assigned AI’s AI work computer, and SU is charged for the time it is used.
- Generate video shows an estimated SU before it starts. The actual charge is for the length of the video produced, and nothing is charged if it can’t be made.
- When a team gets the video editor for the first time, the assigned AI’s AI work computer spec is raised to XL automatically. Narration voices need more memory. More SU is charged while the computer is on, and a single line in the team chat tells you about the change. You can change it any time under AI work computer spec in the assigned AI’s settings. Once you lower it, it isn’t raised again automatically, and it isn’t set back when video work ends.
- Uploaded videos, audio and photos, and generated videos, count toward your account storage and stay until you delete them. Deleting a source frees the space.
Your assigned AI follows up
Video generation, narration and export started by the assigned AI take a few minutes. When one finishes, the assigned AI picks it back up on its own and tells you the result. For work you start yourself, follow progress on the screen.
Good to know
- You can upload mp4, mov or webm video (mp3, m4a, wav or webm for audio), up to 200 MB per file and 3 minutes long.
- Narration and export need an assigned AI.
Advanced
This is not needed for everyday use. Read it only if you want the finer rules for limits and the automatic spec increase.
- The automatic spec increase happens once per assigned AI on a team. It isn’t applied if the spec is already XL or higher or can’t be provisioned, and it never overwrites a value a person lowered.
- Narration can only be created when the AI work computer spec is XL or higher. While the spec is changing, it waits with “Raising the AI work computer spec” and continues afterward.
- Photos are tidied once on upload (rotation applied, colors matched, transparent areas filled with black, long edge reduced to 4096 px or less, animated images use only the first frame). A photo must be 50 megapixels or less.
- One edit holds at most 200 clips.