What happened

Mistral has released Mistral Large 4, nicknamed "Le Chonk", its largest open-weight model to date, published on Hugging Face as mistralai/Mistral-Large-4-1T-A52B. It currently sits at No. 14 on the trending models list with 1,208 likes and 278 recent downloads. The model is a natively multimodal Mixture-of-Experts model with roughly 1 trillion total parameters, taking text and image input and producing text output.

Per the README, it is 「our largest and most capable model to date」, trained from scratch on 3,800 NVIDIA Grace Blackwell GPUs in Mistral's own datacenters in Europe. The model is also served on the Mistral API as mistral-large-4, with an official demo gallery, "Le Chonk Gallery", offering a small shared free budget per visitor during the first week after release.

Key details

• MoE architecture: 128 routed experts (4 active per token) plus 1 shared expert; ~1T total parameters with ~49B active per token (~52B including embeddings and output layers).
• Context: supports a 1M context window for long-context workloads.
• Multimodal: a Pixtral-style vision encoder handling variable image sizes and aspect ratios, able to analyze images, charts, technical drawings, and documents.
• Reasoning mode: three reasoning efforts — low, medium, high (default high) — so one set of weights serves both quick chat replies and complex agentic runs; includes function calling, JSON output, and strict JSON schema enforcement.
• Multilingual: the official language list covers 40 languages; the README claims native fluency in 160+ languages, including every official language of the EU.
• Recommended sampling: temperature=1.0, top_p=0.95.
• Ecosystem support: transformers, vLLM, SGLang, llama.cpp; a separate NVFP4 quantized checkpoint targets Blackwell GPUs with a smaller memory footprint, faster inference, and what the README calls the same performance.
• License: the Mistral Ecosystem Open-Weight License, a modified MIT license for commercial and non-commercial use — but operating a Model-as-a-Service business requires a separate commercial license with Mistral.

Why it matters

On benchmarks, Mistral Large 4 scores 48% on the Artificial Analysis Coding Agent Index, effectively tied with the leading open-weight model GLM-5.3 while using 47% fewer tokens per task on average. The README frames this as 「delivers frontier coding performance with significantly fewer tokens」, implying coding agents are cheaper to run at scale. On the Artificial Analysis Intelligence Index v5 it scores 26, the highest among western open-weight models; the README acknowledges that leading open-weight models from China score higher on that aggregate index, but says ML4 is particularly strong on coding and cybersecurity.

The model is also claimed to 「surpass frontier closed models on visual grounding」 and to be state of the art among open-weight models for cybersecurity work while refusing malicious requests. The checkpoint ships with an embedded DSpark speculative decoding draft head, and Mistral recommends its own coding agent, Mistral Vibe, for building demos.

Who it affects

• Enterprises and teams self-hosting large models: the 1T-parameter, 52B-active MoE with 1M context suits long-document, multimodal, and agentic workloads, though deployment cost and hardware demands are significant.
• Developers of coding agents and tooling: native function calling, JSON schema enforcement, and configurable reasoning effort target agentic workflows directly.
• Multilingual application builders: 160+ languages including all official EU languages.
• Commercial users should note the licensing boundary: direct use falls under the modified MIT license, but offering the model as a service requires a separate agreement with Mistral.

Hugging Face 热门模型第 14 名：mistralai/Mistral-Large-4-1T-A52B，1208 赞，近期下载 278 次。 标签：transformers, mistral_large4, image-text-to-text, vLLM, SGLang, llama.cpp, conversational, en, fr, de, es, pt

README：

Mistral Large 4 - Le Chonk

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

Mistral Large 4 unifies instruction following, reasoning, coding, and agentic workflows in a single set of weights. While a generalist model, it shows especially good performance on workloads such as cybersecurity, finance, and law, and it surpasses frontier closed models on visual grounding. It is natively fluent in 160+ languages, including every official language of the European Union.

Find more information on our blog. The model is also served on the Mistral API as mistral-large-4.

[!Note] The checkpoint ships with an embedded DSpark speculative decoding draft head — see Speculative decoding (DSpark).

[!Note] A quantized NVFP4 checkpoint is also available, with a smaller memory footprint, faster inference and same performance. It targets Blackwell GPUs and supports vLLM and SGLang.

Key Features

Mistral Large 4 includes the following architectural choices:

• ~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.
• Multimodal input: accepts text and image input, with text output. Pixtral-style vision encoder trained to handle variable image sizes and aspect ratios.
• Instruct and Reasoning functionalities with function calls, reasoning effort configurable per request.

Mistral Large 4 offers the following capabilities:

• Reasoning Mode: three reasoning efforts — low, medium, high (default) — so the same model can answer a quick chat reply or work through a complex agentic run.
• Vision: analyzes images, charts, technical drawings, and documents, with state-of-the-art visual grounding among open weights.
• Multilingual: natively fluent in 160+ languages, including all official languages of the European Union.
• System Prompt: strong adherence and support for system prompts.
• Agentic: native function calling, JSON output, and strict JSON schema enforcement.
• Cybersecurity: state of the art among open-weight models for security work, while refusing malicious requests.
• Large Context Window: supports a 1M context window.

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 — operating a Model-as-a-Service business requires a separate commercial license with Mistral.

Recommended Settings

• Reasoning Effort:
  • 'low' / 'medium' / 'high' — default is 'high'.
  • Use reasoning_effort="high" for complex prompts and agentic usage, 'low' for fast chat replies and 'medium' for day to day usage.
• Sampling: for all reasoning efforts, we recommend the following sampling parameters:
  • temperature=1.0
  • top_p=0.95
• System Prompt: define a clear environment and use case, including guidance on how to effectively leverage tools in agentic systems.
• Tools: keep the set of tools well-defined and limit their number to the minimum required for the use case.
• Context: the checkpoint supports a 1M context window.

Benchmarks

We compared Mistral Large 4 with recent open-weight models. We report results on agentic coding and global intelligence here; for more benchmarks, especially on cybersecurity, finance and law, see our blog post.

Agentic Coding

Mistral Large 4 delivers frontier coding performance with significantly fewer tokens, making sophisticated coding agents more economical to run at scale.

On the Artificial Analysis Coding Agent Index, ML4 scores 48%, effectively tied with the leading open-weight model, GLM-5.3, while using 47% fewer tokens per task on average.

Global Intelligence

On the Artificial Analysis Intelligence Index v5, Mistral Large 4 scores 26, the highest among western open-weight models. The leading open-weight models from China score higher on this aggregate index, but Mistral Large 4 is particularly strong on coding and cybersecurity, as shown in our blog post.

Demo Gallery

Le Chonk Gallery is a live gallery of demo Spaces built with Mistral Large 4, from us and from the community: coding, vision and grounding, tool calls, JSON schema, multilingual, and more. Any Space tagged Le_chonk appears there.

Official demos run on the Mistral API with a small shared free budget per visitor for the first week of the release, so you can try the model without paying. You can also use your own Mistral API key.

To build a demo, we recommend Mistral Vibe, our coding agent powered by Mistral Large 4.

Usage

You can find Mistral Large 4 support on multiple libraries for inference and fine-tuning.

We thank all the ecosystem folks who helped us throughout this release to integrate Le

The original belongs to Hugging Face 热门模型 and its authors and is reproduced here for reading. To have it removed, write to support@aisa.one.