
6/20/2024 · Mickie Betz, Ben Krebsbach, Taylor Smith
What this post added
Introduced AI-generated captions for on-demand videos and live stream recordings within Cloudflare Stream, leveraging Workers AI and the Whisper model. This involved pre-processing audio into 30-second batches compatible with Whisper, parallelizing inference requests to Workers AI, and re-ordering out-of-order responses to ensure sequential and accurate WebVTT caption file generation.