Skip to main content
Video Mode turns Foundry’s chat into a generative video studio. Brief in, script and frames come out, edits happen in-chat, the final cut renders. Hit / in the composer and pick Video to enter the mode.
Screenshot coming. Video Mode flow.

​
How a generation flows

  1. Brief. Describe the video — “a 20-second product teaser for a new pair of running shoes, urban skate-park feel”.
  2. Generate. Foundry drafts a script, picks a style, and starts rendering frames.
  3. Polish. Reply with edits — “swap the second shot for a low-angle close-up”, “slower cuts in the middle”. Foundry re-renders only what changed.
  4. Publish. Download the cut or push to a connected destination.
Each stage lives inline in the chat — same message thread, same history. Re-open a video chat tomorrow and the brief, the script, and every render are still there.

​
Tips for briefs

  • Length first. Tell Foundry the target duration in seconds.
  • One mood, one subject. Mixed moods produce mixed cuts. Keep it tight.
  • Reference if you can. “Apple-launch-film style”, “BBC nature-doc style”, “Tarkovsky long-shot” — named references converge faster.
  • Edit in plain language. “hold the establishing shot 1 second longer” works.

​
Limits

  • Generation time scales with duration and quality — expect minutes, not seconds, for full HD output.
  • No upload-and-edit workflow yet. Video Mode is currently generation-first.
  • Surface and exact model lineup are evolving fast; behaviour may shift week-to-week.

Image Mode

Stills with the same engine family.

3D Website Mode

Beyond video — into interactive worlds.