Voxtral 3 Release - Multilingual TTS - 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. Mean Judge Score (1–5).
Metric: Mean Judge Score (1–5). Source: mistral.ai. Saturation forecast: Rough model projection: around 2026. 10 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Eleven v4 | 3.17 | |
| 2 | Voxtral Mini TTS 3.0 | 3.16 | |
| 3 | Eleven v4 Turbo | 3.15 | |
| 4 | Cartesia Sonic 3.6 | 2.95 | |
| 5 | Fish Drama 3 | 2.87 | |
| 6 | Gemini 3.8 Flash Lite TTS | 2.87 | |
| 7 | Gemini 3.8 Flash TTS | 2.83 | |
| 8 | Inworld TTS-2 | 2.63 | |
| 9 | Inworld TTS-2 Flash | 2.52 | |
| 10 | OpenAI gpt-4o-mini-tts | 2.23 |
Interactive version: theaggregate.ai/benchmark?slug=voxtral-3-release-multilingual-tts-overall · How It Works · Data refreshed daily, snapshot 2026-10-21.