Medical STT Benchmark: leaderboard
Speech-to-text benchmark for long-form medical dialogue, ranking cloud and local transcription systems with Medical Word Error Rate on the PriMock57 dataset.
Metric: WER (lower is better). Source: omi.health. Source: github.com. Saturation forecast: Around July 2027. 42 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Pro | 0.08 | #152 |
| 2 | Gemini 3 Pro (Preview) | 0.08 | #66 |
| 3 | Gemini 2.5 Flash | 0.09 | #245 |
| 4 | Gemini 3 Flash (Preview) | 0.11 | #82 |
| 5 | Whisper Large-v3 | 0.12 | #1271 |
| 6 | gemma-4-E4B-it | 0.16 | #566 |
| 7 | Gemma 4 E2B (IT) | 0.19 | #781 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=medical-stt-benchmark · How It Works · Data refreshed daily, snapshot 2026-10-20.