Video Studio
Edit footage, add motion & render
On it. Running Video Studio now and I'll bring the results straight back here.
Try asking
About
Video Studio turns raw camera footage into publish-ready videos. It handles the entire post-production pipeline: silence removal, dynamic zoom cuts, color correction, audio mastering, captions from transcription, motion graphics, B-roll insertion, and final export — all from a single automated workflow.
This is the tool for talking-head content, podcasts, YouTube videos, and online courses. You feed it raw footage and describe what you want; it produces a polished cut without requiring manual editing software or frame-by-frame review.
What you can do
- edit_footage — process raw footage with silence removal, dynamic zoom, color correction, and audio mastering (-16 LUFS); supports two-camera setups with auto sync
- transcribe_video — word-level transcription with frame-accurate timestamps for caption generation
- create_animations — motion graphics and captions built from the transcript, with brand styling
- finalize_video — merge the cut, captions, B-roll, and animations into a final export
- render_video — render a video from a JSON template directly
- render_still — extract a single frame from a video or template
Who it's for
Content creators editing talking-head YouTube videos, podcast producers turning recordings into visual content, course creators who need polished lecture videos, and agencies delivering edited content at scale.
How to use it
- Use list_capabilities first to check available fonts, transitions, and output formats
- Run edit_footage with your raw video URL — it removes silences, applies zoom cuts, and masters the audio
- Run transcribe_video to get word-level timestamps for caption-accurate animations
- Run create_animations to add captions and motion graphics in your brand style
- Run finalize_video to merge everything into the final output
Getting started
Upload your raw footage to a public URL and pass it to edit_footage. All async skills return a job ID — the tool will poll automatically until the result is ready.
Edit raw footage into a professional cut
Analyze edited video and generate branded motion graphics for key…
Assemble the final video by combining edited footage with animation…
Transcribe a video with word-level timestamps using speech-to-text
Edit raw footage into a professional cut. 1-2 cameras with auto audio sync, silence removal, dynamic zoom (normal/punched-in/tight), color correction, and audio mastering (-16 LUFS) with EQ, compression, and de-essing via ffmpeg.
Analyze edited video and generate branded motion graphics for key moments. Detects verbal cues, data points, comparisons, and processes, then renders overlays (counters, flowcharts, quotes, lists, highlights) via Remotion. Returns timestamped animation files for finalize_video.
Assemble the final video by combining edited footage with animation overlays and B-roll clips. Outputs high-quality MP4 (libx264, 320k AAC). Use after edit_footage and create_animations.
Transcribe a video with word-level timestamps using speech-to-text. Returns the full transcript text, time-coded segments, and per-word timestamps. Useful standalone for subtitles, content analysis, or as a precursor to create_animations. Supports multiple languages.
Render a video from an ordered array of scenes. Each scene has positioned elements (text, images, shapes) with full layout, animation, and styling control. Use template_props for quick presets. Returns a downloadable MP4. Call list_capabilities first.
Render a single scene as a still image (PNG, JPEG, or WebP). Same element-based canvas as render_video. Useful for thumbnails, social graphics, and previews.
List all available capabilities: footage editing pipeline (edit, animate, finalize), scene-based video creation (elements, animations, templates), transcription, and all configurable options. Call this first to understand the full toolkit.
- Added subtitle, expanded description, and agent instructions
- Added edit_footage skill: multi-camera sync, silence removal, zoom levels, color correction, audio mastering
- Added create_animations skill: transcription-based motion graphics with brand styling
- Added finalize_video skill: animation overlay, B-roll insertion, final encode
- Added transcribe_video skill: word-level transcription
- Renamed tool to Video Studio to reflect full production pipeline
- Initial release
Powered by
Related
Guides
Questions
What is the best starting point for a talking-head edit?
Start with `list_capabilities`, then use `edit_footage` for silence removal, dynamic zooms, color correction, and audio mastering before you add motion graphics.
Can it turn a transcript into captions or motion graphics?
Yes. `transcribe_video` gives you word-level timing, and `create_animations` uses those cues to build overlays like counters, quotes, lists, and flowcharts.
How do I finish the edit with B-roll and overlays?
Use `finalize_video` after the base edit and animation pass. That is the step that assembles the final export with B-roll clips, overlays, and the chosen output encoding.
Does it support multi-camera footage?
Yes. `edit_footage` supports one or two cameras with automatic audio sync, so it works for simple interviews as well as single-camera content.