Mistral Large 4 Preview Benchmarks Just Dropped
Informal-Trouble2183 · reddit · 2026-10-06
A Reddit user shared the just-released benchmark results for the Mistral Large 4 preview (codename le chonk), consistent with Mistral's own update: the RL run is still ongoing with continued improvements, and the full release is expected later this month. Benchmark charts are in the post.
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