Voxtral 3 TTS Scores 84.4% on Pronunciation Robustness, Ranks #13 of 32 Models
ArtificialAnlys · x · 2026-10-21
On Artificial Analysis's Pronunciation Robustness benchmark, Voxtral 3 TTS scores 84.4% overall, ranking #13 of 32 models. The benchmark tests whether TTS models pronounce challenging text correctly, with human reviewers judging clips against pre-agreed pronunciations:
- Standalone Terms: 93.9%, #3 behind Fish Audio Drama 3 Preview (95.9%) and Eleven v3 Conversational (95.1%)
- Contextually Appropriate: 91.9%; Gemini 3.8 Flash TTS leads the category at 97.9%
- Expanding Shorthand: 80.2%, level with Qwen-Audio-3.1-TTS-Plus; Ming-3.0-tts leads at 96.5%
- Preserving Exact Sequences: 75.5%; Ming-3.0-tts tops the category at 90.6%
In the Controlled Voice Arena, Voxtral 3 TTS scores 1,068 (#12), slipping to #20 in Entertainment; by accent it ranks #10 in UK English (1,071) and #17 in US English (1,056).
More from Models
- Ternary-Bonsai-2-9B: 2-bit ternary vision-language model trends on Hugging Face for on-device use — prism-ml · 2026-10-21
- Grok Bot DAU hits new peak as users report it books flights and pays bills — elonmusk · 2026-10-21
- Interconnects: China proposes export pre-approval for frontier models as DHS bars DeepSeek from federal procurement — natolambert · 2026-10-21
- Researcher urges the community to stop using PickScore as a reward model, recommends HPSv3 instead — gowthami_s · 2026-10-21
- OpenAI boosts COBOL, Fortran and Visual Basic support to modernize legacy enterprise systems — CurieuxExplorer · 2026-10-21
- NVIDIA Hosts Ask-the-Experts Session on Nemotron 3.5 Super VL — NVIDIAAI · 2026-10-21