Agentic WeeklyGet the weekly

Reported News

Mistral releases the weights of Mistral Large 4 (Le Chonk), a 1T-parameter multimodal MoE model

On Oct 29, 2026 Mistral released the weights of an improved Mistral Large 4 (Le Chonk), a multimodal mixture-of-experts model with about 1T total parameters and a 1M-token context, under the Mistral Ecosystem Open-Weight License. Mistral calls it the best open-weight model for coding.

By Agentic Weekly · Evidence: Reported · How we work

Changes reported
6
listed below, each from a source we read
We tested
0
we did not run it ourselves
Gaps we flag
4
things we could not verify
Sources
2
2 primary, linked below

What changed

Changes the sources report, with the evidence status of each
#What the source saysStatus
01Mistral released the weights of Mistral Large 4 (Le Chonk), three weeks after the model entered public preview. Mistral says the released model is improved by further reinforcement learning, mainly in coding and defensive cybersecurity.Reported
02Hugging Face describes it as a multimodal Mixture-of-Experts model that takes text and images and outputs text, with about 1T total parameters and a 1M-token context, shipped as FP8 and NVFP4 checkpoints under the Mistral Ecosystem Open-Weight License.Reported
03Hugging Face says the license is an MIT derivative that sets limitations for model-as-a-service offerings, and that the model has day-0 support in transformers, vLLM and llama.cpp, plus TRL fine-tuning, Inference Endpoints and GGUF quants.Reported
04The model now has three reasoning modes, low, medium and high (the default); Mistral says low replaces the old None mode, and Hugging Face notes reasoning cannot be turned off.Reported
05Mistral's own benchmark claims: 19.7% on Cognition's private FrontierCode 2.0, 41% on Terminal-Bench 4.0 (up 13 points from the preview), 48% on the Artificial Analysis Coding Agent Index and 51% on the Artificial Analysis Cyber Index.Reported
06Mistral says a specialized cybersecurity variant for penetration testing and red-teaming is in preview only for vetted enterprise design partners, and that it is tightening guardrails after seeing potential uplift for real attacks.Reported

reported by the sourcemeasured or run by usnot verified

Primary source: Mistral AI: Mistral Large 4 (Le Chonk) is now open weight. We are the fast, independent index; the vendor’s own page is the source of record for this release.

In the source’s words

“Today, we're releasing the weights of this improved Mistral Large 4.”

“It’s distributed under the Mistral Ecosystem Open-Weight License (an MIT derivative that establishes limitations for model-as-a-service offerings).”

“Though we observed the model refusing to comply with black-box attacks against live targets, we see a potential for uplift to conduct real attacks.”

What we have not verified

  • We did not download or run the model; this brief is Reported from Mistral's post and the Hugging Face blog.
  • All benchmark numbers are vendor claims we did not reproduce. The sources also frame cyber results differently: Mistral cites 51% on the Artificial Analysis Cyber Index, while Hugging Face cites 82% on its reproduce-and-patch task.
  • Active parameter counts differ by source: Hugging Face says 49B active, while the checkpoint name (A52B) and Mistral's preview docs say 52B. We did not read the full license text.
  • Neither post says whether API pricing or the API's preview status changed with the weights release.

This brief is Reported: written from the sources below, not from running the release ourselves.

Sources

Spotted an error? See our corrections policy or email the address on the contact page. Markdown version for agents: /news/mistral-large-4-open-weights.md

Want these briefs in your Google Top Stories? Add Agentic Weekly as a preferred source on Google.