Blogger blind-tests Mistral Voxtral 3 French TTS against ElevenLabs after suspicious internal benchmarks
emmanuelvivier · x · 2026-10-22
Mistral released Voxtral 3 claiming to beat ElevenLabs and OpenAI, but its benchmarks are mostly internal LLM-scored with no public French rankings. Blogger leploutos ran his own French blind test — three sentence types (customer-service numbers, trap phrases with addresses/brands/€ amounts, emotional lines), matched voices, normalized volume, no emotion tags. He also flags a gotcha: Mistral's docs list only one legacy French voice, while the API actually offers four new Voxtral 3 voices (Marie, Manon, Léa, Camille).
More from Models
- Recursive AI ships Flint, a 40B-A2.5B MoE claimed to match Qwen3.6-27B, weights on Hugging Face — ChengleiSi · 2026-10-22
- DeepSeek ships V4.1-Flash-1017, now the #1 open model on RSI-Inference Bench — scaling01 · 2026-10-22
- OpenAI's Cybersecurity Guardrails Block a Dev From Auditing His Own Patched Server — Next_Technology6361 · 2026-10-22
- TPN-006 hits Hugging Face: 7.7GB compressed model beats its 17GB Mistral teacher — markjeffrey · 2026-10-22
- Kimi K3.1's latest checkpoint draws 49k-token SVG portrait zero-shot in 22 minutes — lyraxana · 2026-10-22
- MiniMax launches M3.1 Flash at 1/3 the cost and 3x the speed of M3, weights out Oct 23 — MiniMax_AI · 2026-10-22