Mistral Launches Large 4: 1T-Parameter Multimodal Model With 49B Active, Weights Soon
GuillaumeLample · x · 2026-10-06
- Mistral released Mistral Large 4 ("Le Chonk"): 1T total parameters, natively multimodal, only 49B active. Preview API live today; open-weights release expected end of October.
- Claims: best open-weights model from the US or Europe on aggregated benchmarks; SOTA on cyber defense, manufacturing and finance workloads; beats closed frontier models on visual grounding. Built end-to-end in Europe, deployable via Mistral's own cloud.
- Researcher color (AlbertQJiang): good in actual usage (used it to study Lacan and Merleau-Ponty); DeepSWE performance over 60 won a tattoo bet; hints an "even chonkier" model is currently training on trillions of tokens.
Related event: Mistral Unveils Mistral Large 4, a 1T-Parameter Open Multimodal Model(42 posts)→
More from Models
- Google rolls out Nano Banana 2.1 with 4K output at $0.076 per image — testingcatalog · 2026-10-06
- Space Bunny Alpha lingers on OpenCode a few days longer than OpenRouter — sull · 2026-10-06
- Analysis: US Models Are Now Cheaper per Task Than Chinese Ones, Upending a Widely Held Assumption — sebkrier · 2026-10-06
- GPT-6.1 Sol spotted in Copilot 365 Premium, ChatGPT rollout likely next — koltregaskes · 2026-10-06
- Are models only improving at verifiable domains? AI stories winning prizes spark debate — erikphoel · 2026-10-06
- Reflection AI's upcoming open-weight model could bring the open-source crown back to the US — The AI Daily Brief · 2026-10-06