Voxtral 3 Release - Realtime FLEURS 35 - Overall: leaderboard
Voxtral 3 release evaluation (2026-10-20): exact chart marks retain model, series and score. FLEURS 45 macro-averages WER, using CER for Japanese, Korean, Thai and Chinese; language forced/inferred runs stay separate. Diarization averages cpWER over AMI-IHM, AMI-SDM, Switchboard, CallHome and AISHELL-4 with 30-minute clips. Context-bias recall uses Earnings22. Realtime latency configurations remain distinct. TTS tests retain their release-specific populations and judge; chart x-axis prices are never scores. WER/CER (%).
Metric: WER/CER (%). Source: mistral.ai. Saturation forecast: Rough model projection: around 2026. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Voxtral Mini Transcribe 3 | 4 | |
| 2 | Gemini 3.5 Realtime Transcribe | 4.2 | |
| 3 | Voxtral Mini Realtime 3 · 2400 ms | 5.7 | |
| 4 | Microsoft AI Transcribe 2 Streaming | 5.8 | |
| 5 | Voxtral Mini Realtime 3 · 960 ms | 6 | |
| 6 | Voxtral Mini Realtime 3 · 480 ms | 6.4 | |
| 7 | Voxtral Mini Realtime 3 · 240 ms | 7.4 | |
| 8 | GPT Live Transcribe | 8.4 | |
| 9 | Scribe v2 Realtime | 8.7 |
Interactive version: theaggregate.ai/benchmark?slug=voxtral-3-release-realtime-fleurs-35-overall · How It Works · Data refreshed daily, snapshot 2026-10-21.