Mistral releases Large 4 weights, but paid hosting needs a separate licence

Mistral says it is releasing the weights of Mistral Large 4, which it calls its flagship 1-trillion-parameter, natively multimodal model, an open-weight hybrid instruct-and-reasoning mixture-of-experts model.

Earlier this month, we released Mistral Large 4 (Le Chonk), our flagship 1-trillion-parameter, natively multimodal model, in public preview. ... Today, we're releasing the weights of this improved Mistral Large 4. ... Open-weight hybrid instruct-and-reasoning MoE with multimodal input

Mistral's post says use, modification, fine-tuning and commercialisation are allowed, but businesses that operate a Model-as-a-Service offering must obtain a separate commercial license from Mistral.

The model is free to use, modify, fine-tune, and commercialize, with a single condition: businesses that operate a Model-as-a-Service offering must obtain a separate commercial license from Mistral before using the model commercially.

Mistral's post lists API prices of 0.70 US dollars per million input tokens, 0.07 dollars for cached input and 2.10 dollars per million output tokens.

Starting today, Mistral Large 4 is priced at $0.70 input / M, $0.07 cached input / M, and $2.10 output / M tokens.

Mistral says Large 4 scores 48% on the Artificial Analysis Coding Agent Index, effectively tied with GLM-5.3, using 47% fewer tokens per task on average.

On the Artificial Analysis Coding Agent Index, Mistral Large 4 scores 48%, effectively tied with the leading open-weight model, GLM-5.3, while using 47% fewer tokens per task on average.

Mistral says it measures an average refusal rate of 96.6% on malicious cyber requests, the highest it measures among open models.

while achieving an average refusal rate of 96.6% on malicious cyber requests, the highest we measure among open models.

Mistral says that, from monitoring cybersecurity-related use in its own live traffic since the preview, it sees a potential for uplift to conduct real attacks, and that it is modifying its guardrails accordingly.

In the time since the preview, we carefully monitored cybersecurity-related use in our live traffic. 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. We are modifying our guardrails accordingly, and will continue updating them in the months to come.

Mistral says a specialised cybersecurity variant is in preview for vetted design partners such as Orange Cyberdefense.

We are also providing enterprise design partners such as Orange Cyberdefense, with preview access to a specialized variant of Mistral Large 4 optimized for advanced cybersecurity workflows ... Access is subject to rigorous vetting and contractual safeguards

The model card describes Mistral Large 4 as a 1 trillion-parameter natively multimodal Mixture-of-Experts model with 49B active parameters per token (52B including embeddings and output layers).

Mistral Large 4 is our largest and most capable model to date: a 1 trillion-parameter natively multimodal Mixture-of-Experts model with 49 billion active parameters per token (52 billion including embeddings and output layers)

The model card lists a 1M-token context length and a granular Mixture-of-Experts design with 128 routed experts, 4 active per token, and 1 shared expert.

~1T total / ~52B active parameters: granular Mixture-of-Experts with 128 routed experts (4 active per token) and 1 shared expert. 1M context length, for long-context workloads.

The licence is the Mistral Ecosystem Open-Weight License, which the model card describes as a modified version of the MIT license for both commercial and non-commercial use.

We release this model under the Mistral Ecosystem Open-Weight License: a modified version of the MIT license, for both commercial and non-commercial use

The model card says the model was trained from scratch on 3,800 NVIDIA Grace Blackwell GPUs in Mistral's European datacenters.

trained from scratch on 3,800 NVIDIA Grace Blackwell GPUs in Mistral's own datacenters in Europe.

Mistral's company page says its co-founders saw the need for a different kind of AI company: a European leader.