Mistral Opens Public Preview Of One-Trillion-Parameter Large 4
· TexaHub · LLMs & Foundation Models, Regional Tech Ecosystems
Mistral AI released Mistral Large 4, a multimodal mixture-of-experts model trained on 3,800 Nvidia Grace Blackwell GPUs in Europe, with open weights promised by the end of October.
Specifications and pricing
Mistral AI put Mistral Large 4 into public preview on 6 October 2026. The company describes it as a natively multimodal, hybrid instruct-and-reasoning mixture-of-experts model with about 1 trillion total and 49 billion active parameters, fluent in more than 160 languages [1][2].
It was trained on 3,800 Nvidia Grace Blackwell GPUs in Mistral's European data centers. API pricing is $1.36 per million input tokens and $4.18 per million output tokens [1][2].
Benchmarks and weights
Mistral reports 61.7% on DeepSWE v1.1 and 82% on a vulnerability reproduction and patching test, and says the model ranks among the top five on the Artificial Analysis Cyber Index [1]. MLQ.ai reported a score of 38 on the Artificial Analysis Intelligence Index, against 58 for Claude Opus 5.5 [2]; these are vendor-cited and third-party figures not independently checked here. Weights are due at the end of October, with MLQ.ai citing 27 October after red-teaming [1][2].
Sources
[1] Mistral Large 4 — Mistral AI, 6 Oct 2026
[2] Mistral Launches Large 4 Preview, a 1 Trillion-Parameter AI Model — MLQ.ai, 6 Oct 2026