AuralKit 3.0 adds transcription of selected audio tracks from local video with timestamps
rudrank · x · 2026-11-13
The open-source Swift package AuralKit 3.0 can now transcribe a chosen audio track from a local video or media file on iOS and macOS 27, with results preserving timestamps on the media timeline. stopTranscribing() is replaced by finishTranscribing() (finish capture and get final results for queued audio) and cancelTranscribing() (abandon the run; the stream ends with CancellationError). On OS 27, screen capture can feed compatible sample buffers directly into Speech, buffered audio drains on Finish, and the microphone input uses a throwing, Sendable tap. Fixes contributed by Alex Ehlke.
Related event: AuralKit 3.0 Adds Audio Track Transcription for Local Videos(2 posts)→
More from coding & agent
- Free paper-chat tool with MCP tools now open for agents — omarsar0 · 2026-11-14
- ZigZag: open-source embroidery editor with MCP lets AI agents design stitches — simpsoka · 2026-11-14
- Google Antigravity adds Gemini 4 Hydrogen and Orchestrator for long-horizon agent tasks — minchoi · 2026-11-14
- DAIR.AI launches daily-updated paper collections: 825 harness engineering, 540 recursive self-improvement papers — omarsar0 · 2026-11-14
- When Claude went down, Orchestrator users one-clicked their conversations over to Codex — julianweisser · 2026-11-14
- ACE-RL: RL-trained agents learn to curate experience playbooks that boost future rollouts — kaist-ai · 2026-11-14