Open models transcribe 8h38m of mock GP consultations in 18m38s on a single GPU — 27.8x faster
MaziyarPanahi · x · 2026-11-19
Maziyar Panahi demoed transcribing and drafting notes for 57 mock GP consultations (PriMock57 dataset, 8h38m of audio) in 18m38s on a single GPU using only open models — 27.8x faster than real time, with each draft ready 20s after its call.
He clarified the calls are mock consultations with clinicians playing patients (CC BY 4.0), not real patients, and every draft is for doctor review and sign-off. Many doctors still write notes late into the night, which this workflow aims to relieve.
More from Apps
- Flowith 3.0 launches: orchestrate SOTA models on one canvas, $9.99 buys 600 videos — lxfater · 2026-11-19
- Synthesia integrates with Claude: generate and publish AI videos from chat — synthesiaIO · 2026-11-19
- Details emerge on Gemini Notebook's Writing Partner, Coach and No AI modes — testingcatalog · 2026-11-19
- Google reportedly prepping three Writing Modes for Gemini Notebook — testingcatalog · 2026-11-19
- CapCut gutted image-to-image credits: free tier once allowed dozens daily, now even Pro runs dry — Dyzow5076 · 2026-11-19
- Meta's Muse AI Read Encrypted Signal Messages, Internal Chats Reveal Extreme Privacy Violations — Justgototheeffinmoon · 2026-11-19