FlowFrame
Training content shouldn't require a video production team.
1 / 3
Step-based, not screen-based
Screencasts break the moment someone moves a button. Written SOPs go stale the day they're published. FlowFrame splits instructional content into atomic steps — each with a visual, narration, and instruction. When something changes, you update the step, not the whole video.
- Local-first — your data stays on your machine
- Show / Say / See per step
- Exports to MP4, GIF, or structured HTML
AI that builds the demo for you
Describe what you want to show and an agent writes the run-of-show, sources visuals, generates narration, and exports a finished video. You stay in the director's chair.
Platform macOS desktop app
Stack Electron, React, FFmpeg
License MIT / open source
FlowFrame is in alpha — dedicated landing page coming soon.
Get early access →