Mistral Large 4: Release, Price and Benchmarks

Mistral Large 4 is Mistral AI's new flagship model, nicknamed "le Chonk." It has been in public preview through the Mistral API since October 6, 2026, but the open weights are still not out: as of October 23, Mistral's Hugging Face page is a countdown with an expected release of October 31, 2026 and no files to download. The two week launch discount has also ended, so the API now costs the full list price of $1.36 per million input tokens and $4.18 per million output tokens.

This guide covers what Mistral has confirmed: the model's size, how to get access, the current price, the benchmark results Mistral reports, and what is still unknown. Numbers come from Mistral's launch post, its Large 4 model card and its changelog. Benchmark figures are Mistral's own claims and have not been independently checked. The page reflects the situation on October 23, 2026.

What changed since launch (October 23 update)

• The launch discount is over. Mistral's changelog entry for October 6 says "Launch pricing: 50% off for 2 weeks." The model card now shows only the list price, so the half price rate of $0.68 and $2.09 no longer applies.
• The weights date still holds. In an October 21 announcement with NVIDIA, Mistral repeated that the weights are "planned for release at the end of October." The Hugging Face placeholder still lists October 31 as the current ETA.
• No license yet. Neither the model card nor the Hugging Face page names a license. The model card only says "Open."
• It is on OpenRouter. OpenRouter's public model list includes Mistral Large 4 as mistralai/mistral-large-4-0 at the same per token prices as Mistral's own API and a context window of 1,048,576 tokens.
• A new NVIDIA project builds on it. Mistral and NVIDIA will use Large 4 for engineering simulation work and are developing a frontier base model through the NVIDIA Nemotron Coalition.

What Mistral Large 4 is

Large 4 is a mixture of experts model. Mistral's model card lists 1.05 trillion total parameters and 52 billion active per token, plus a separate 1.6 billion parameter vision encoder. The Hugging Face page explains the other figure you will see quoted: 49 billion parameters are active per token, or 52 billion counting the embedding and output layers. The context window is 1 million tokens.

It is natively multimodal (images as well as text in) and works as a hybrid "instruct and reasoning" model, so one model handles chat, step by step reasoning and agentic tool use. Mistral calls it "our largest and most capable model to date." Other confirmed details:

• Training hardware: trained from scratch on 3,800 NVIDIA Grace Blackwell GPUs in data centers Mistral operates in Europe.
• Languages: training data covers more than 160 languages, including every official EU language.
• Version tag: version 26.10. Its predecessor, Mistral Large 3, was released in December 2025.
• API features: the model card lists structured outputs, function calling, document Q&A, batching and Mistral's agents and built in tools endpoints.
• Still improving: at launch Mistral said the reinforcement learning run behind the preview was "still in flight," so the model may change before the weights ship.

Release date: preview now, weights later

There are two release dates to keep apart:

1. Preview API: available since October 6, 2026. Anyone with a Mistral account can call mistral-large-4 through Mistral Studio, and it is also listed on OpenRouter.
2. Open weights: expected October 31, 2026. Until then, Mistral says it is red teaming the model with cybersecurity leaders, vetted partners and state authorities. Those testers get a version with "reduced moderation and expanded cyber capabilities." Mistral also promises architecture details, more benchmarks and its post training method alongside the weights.

Google's Gemini 4 Argon also went first to vetted cyber defenders. The difference is the destination: Mistral says Large 4 will end up as downloadable weights, while Argon and GPT-6 Astra stay behind an API.

What license will it use?

Mistral calls Large 4 "open weight," but the license is not published. For comparison, Mistral Large 3 shipped under Apache 2.0, and Mistral Medium 3.5 under a modified MIT license. Don't assume any terms until the license text appears with the weights. Moonshot is taking a similar path with Kimi K3.1 Pro, which is on the API now with weights promised by November 4, 2026.

Mistral Large 4 pricing

With the launch discount gone, these are the current API prices on Mistral's model card. OpenRouter lists the same rates.

Read aloud: one dollar thirty six per million input tokens, fourteen cents for cached input, and four dollars eighteen per million output tokens. As a worked example, a job that sends 1 million input tokens and gets 200,000 output tokens back costs about $1.36 plus $0.84, or roughly $2.20 at list price.

These are still preview prices and could change at general availability. Once the weights are out you can self host, and your cost becomes your own hardware or cloud bill. A model this size needs multi GPU servers, not a laptop.

If you want to chat rather than call an API, Mistral's consumer assistant is Le Chat, which Mistral now runs as part of its Vibe app at chat.mistral.ai. Mistral has not announced when Large 4 becomes a model there.

Benchmarks Mistral reports

All figures below are reported by Mistral:

What Mistral says these show:

• Coding: its Coding Agent Index score puts it ahead of DeepSeek V4 Pro and Qwen3.8 Max. In a blind human evaluation run with Surge AI, it ranked second of five models (3.74 out of 5), ahead of Kimi K3 and GLM-5.3 but behind Claude Opus 5 (4.22).
• Cybersecurity: a top five place on the Artificial Analysis Cyber Index, and 82% on a test that asks a model to reproduce and then patch a real vulnerability, which Mistral calls the highest of any model.
• Knowledge work: in legal and financial evaluations run by vals.ai, Mistral says Large 4 beat GPT-6 Astra.

Notice what Mistral does not claim. It never says Large 4 beats the best closed models overall. Its headline claim is that Large 4 is competitive with the strongest open models anywhere and "significantly" ahead of any open weight model built in the US or Europe.

The cybersecurity refusal debate

The most debated part of the launch is refusals. Mistral says several closed models, including Claude Opus 5.5 and GPT-6 Astra, "score near zero" on the vulnerability reproduction test because they refuse the task. Mistral argues that defenders often have to prove a flaw is real before they can fix it.

The same post claims Large 4 refuses malicious cyber prompts more often than any other open model, measured on refusal sets drawn from JailbreakBench, StrongREJECT and AgentHarm. Both can be true. But open weights change the safety picture: once weights are public, anyone can fine tune the safeguards away. Our guide to the Preparedness Framework vs the Responsible Scaling Policy explains how OpenAI and Anthropic decide what is too risky to release. On prompt injection, Mistral cites Lakera's public B3 benchmark, where Large 4 resisted 93.3% of attacks.

Mistral Large 4 vs Opus 5.5, GPT-6 and Gemini

"Mistral Large 4 vs Opus 5.5" and "vs GPT 6" are two of the most common searches about the model. No independent, like for like comparison has been published by any of the labs, so here is what the official material supports:

Pros of Large 4 today: a frontier class model with a firm date for downloadable weights, a 1M token context and a European provider. Cons: no license yet, no independent benchmarks, the preview may still change, and self hosting needs serious hardware.

For the closed flagships, see GPT-6 Astra vs Claude Fable 5.1, Claude Fable vs Opus vs Mythos and our ChatGPT vs Claude vs Gemini vs Grok comparison. For everyday chat, ChatGPT, Claude and Gemini remain the easier picks.

Who should try it now

• Developers comparing models: test the preview on your own prompts now, so you know whether self hosting is worth it on October 31.
• Security teams: the clearest target audience, but the public preview is the standard build, not the reduced moderation version given to vetted partners.
• European organizations with data sovereignty needs: Mistral says it will run a European deployment end to end, under European law.
• Everyday chat users: no need to switch. Wait until Large 4 reaches Mistral's chat app and independent reviews appear.

What to watch next

• The weights and license text, expected October 31, 2026.
• Architecture and post training details, promised with the weights.
• Independent benchmark runs once anyone can download the model.
• Specialized models built on Large 4, which Mistral says are coming.

We'll update this guide when the weights ship.

From Ahzy Capital (our own products, labeled as such): Get the free AI Agent Cheat Sheet: 24 copy and paste AI agent setups for 8 types of small business, with prompts and workflows. We make and sell these, so they aren't independent picks.

FAQ

When is the Mistral Large 4 release date?

The preview API launched on October 6, 2026. The open weights are expected on October 31, 2026, the current ETA on Mistral's Hugging Face page, and Mistral repeated on October 21 that they are planned for the end of October.

How much does Mistral Large 4 cost?

$1.36 per million input tokens, $0.14 per million cached input tokens and $4.18 per million output tokens. The 50% launch discount ran for two weeks after October 6 and has ended. Once the weights ship you can self host and pay only for your own compute.

Is Mistral Large 4 on OpenRouter?

Yes. OpenRouter lists it as mistralai/mistral-large-4-0, at the same prices as Mistral's own API and with a context window of about 1 million tokens.

How big is Mistral Large 4?

1.05 trillion total parameters and 52 billion active (49 billion excluding embedding and output layers), plus a 1.6 billion parameter vision encoder and a 1M token context window.

Is Mistral Large 4 open source?

Mistral calls it open weight, but the weights and license are not out yet. Large 3 used Apache 2.0, but Mistral has not said what Large 4 will use, so read the license when it appears.

Is Mistral Large 4 better than GPT-6 or Opus 5.5?

Not overall, based on Mistral's own claims. It reports narrow wins over GPT-6 Astra on visual grounding and some legal and finance tasks, and it ranked below Claude Opus 5 in Mistral's coding evaluation. No independent head to head exists yet.