GPT-Transcribe halves Whisper’s error rate on real-world audio benchmarks
OpenAIDevs · x · 2026-07-29
OpenAI shared benchmark results for GPT-Transcribe:
- On Context Aware ASR, semantic accuracy improved from 41.6% without free-form context to 45.2% with context.
- On Common Voice across 22 languages, it reached a 19.27% transcription error rate, versus 40.37% for Whisper (whisper-1).
- On Real-World Audio Recording across 9 languages, it reached an 8.98% error rate, versus 15.21% for Whisper (whisper-1).
More from Models
- NVIDIA open-weight model ties pplx-decider at #1 on 669-case clinical benchmark — MaziyarPanahi · 2026-10-16
- Claude Code adds idle compaction to save 1h prompt cache on idle sessions — EricBuess · 2026-10-16
- Thrive Holdings launches real-work AI evals from 75+ businesses, benchmarking 13 models — samaysham · 2026-10-16
- Microsoft's MAI-Transcribe now handles 61% of all transcription on OpenRouter — mustafasuleyman · 2026-10-16
- Reflection AI Releases Open-Weight 501B MoE Beam for Agentic Coding on AWS — AWS ML Blog · 2026-10-16
- Dev RL-tunes Qwen 27B against slop with Meta's XAR method, lands top-4 on EQbench — Mysterious_Hearing14 · 2026-10-16