GoKawiilMistral releases open weights for Large 4 'Le Chonk' with improved coding scores
Mistral has published the full weights for Mistral Large 4, the trillion-parameter model nicknamed 'Le Chonk' that first appeared in preview on Oct. 6. New benchmark results show its score on the Terminal-Bench 4.0 coding test rising from 28% to 41%, putting it slightly ahead of GLM-5.3 on that metric, though GLM-5.3 and Codex GPT-6.1 Terra still edge it out on Artificial Analysis's overall composite index. Mistral also says the model uses 47% fewer tokens per task on average than comparable open models.
GoKawiil's interpretation of the reporting above, not reported fact.
The gains suggest open-weight Western models are closing the performance gap with both leading proprietary systems and Chinese open models like GLM-5.3, which could pressure labs to compete more on efficiency and cost rather than just raw benchmark scores. Mistral's emphasis on cybersecurity testing and data sovereignty also echoes efforts by OpenAI and Anthropic, suggesting model security evaluation is becoming a standard expectation across the industry. However, benchmark scores should be read cautiously, as the report itself notes that models can be optimized to game these tests.
- Mistral Large 4's coding benchmark score jumped from 28% to 41% between preview and full release.
- It now narrowly beats GLM-5.3 on Terminal-Bench 4.0 but trails slightly on Artificial Analysis's overall Coding Agent Index.
- Mistral says the model uses 47% fewer tokens per task than rival open models while emphasizing cybersecurity and data sovereignty.
Source: zdnet.com — Radhika Rajkumar, 2026-10-29
Published there as: “Mistral’s Le Chonk is now a top open model for coding, ranking alongside Codex GPT-6.1 Terra”
Read the original report → The summary and analysis above are GoKawiil's own, written from reporting by the source above. Facts and quotes belong to the original publisher.