Captions in seconds, not hours.
Caption Toolkit: automating subtitles in After Effects.
Context
Creating captions inside After Effects is usually slow and manual; designers have to manually transcribe audio, break lines, time each caption and build text layers one by one.
Tools like CapCut and Premiere Pro automate this, but After Effects doesn't offer anything similar.
Caption Toolkit was built to bring that level of automation directly into After Effects, producing clean, editable captions in seconds instead of hours.
Caption Toolkit incorporated into an existing Vidsy tool.
Role
- Problem identification
- Pipeline architecture
- UX Design
- Full-stack development (Bolt CEP, React, TypeScript, ExtendScript)
Caption Toolkit utilises OpenAI Whisper & Assembly AI to generate captions.
Solution
- Exports timeline audio (all tracks or selected) directly from After Effects for transcription.
- Sends audio to OpenAI Whisper or AssemblyAI depending on the transcription needs.
- Parses the returned SRT and generates fully editable text layers with accurate in/out points across the timeline.
- Delivers the same flexibility as manual captions without any of the manual labour.
Ae timeline generated by the caption toolkit in seconds.
Impact
The API gets called every day. That confirmed adoption before any feedback did.
Caption Toolkit is the only captioning workflow for After Effects at Vidsy. What used to take an hour of manual work now completes in seconds, without anyone leaving the app.
