Mistral 发布了 Mistral Large 4（Le Chonk），这是一个拥有 1 万亿参数的多模态模型，现已提供改进后的编码、网络安全和推理效率。它在开源权重基准测试的编码和防御性安全领域处于领先地位，同时提供三种推理模式以平衡成本与性能。该模型完全开源且在欧洲训练，通过可定制、自托管的部署选项，赋能主权 AI。
本月早些时候，我们以公开预览的形式发布了旗舰级 1 万亿参数原生多模态模型 Mistral Large 4（Le Chonk）。预览阶段的反响超出了我们的预期，我们要感谢开发者社区、早期访问客户和合作伙伴的支持、协作与反馈。
Mistral 正在定义主权 AI 类别并引领开放前沿。我们致力于构建世界上最强大的模型并以开放方式发布，赋予开发者、企业和机构拥有、定制和控制其 AI 的能力。
Mistral Large 4 在主要编码基准测试中凭借强劲表现进入公开预览阶段。仅仅经过三周的额外强化学习，它便跃居开源权重编码的前列，在 FrontierCode 和 Terminal-Bench 上领先，并在 Artificial Analysis Coding Agent Index 上与最佳开源权重模型持平，同时具有显著更高的令牌效率。
在 Cognition 的私有 FrontierCode 2.0 基准测试中（该测试评估端到端代码质量以及存储库维护者是否会合并智能体的更改），Mistral Large 4 以 19.7% 的得分领先。
在 Artificial Analysis Coding Agent Index 中，Mistral Large 4 得分为 48%，与领先的开源权重模型 GLM-5.3 基本持平，同时在每个任务上平均使用的令牌减少了 47%。
这些优势也延伸到了基于终端的编码工作流。在 Terminal-Bench 4.0 上，Mistral Large 4 相比其公开预览版本提升了 13 个百分点，达到 41%，在开源权重模型中位居榜首，同时输出令牌减少了 30%。
除了基准测试外，Mistral Large 4 的编码能力也得到了人类偏好的验证。在 Surge Code 的双盲配对评估中，Mistral Large 4 优于包括 GLM-5.3、Kimi K3 和 DeepSeek V4.1 Flash 在内的主要开源权重模型，在被评估的模型中仅排在 Claude Opus 5.5 之后。
Mistral Large 4 在 Artificial Analysis Cyber Index 中得分为 51%，该指数衡量模型在代码库中发现漏洞、复现并验证漏洞，以及在不断现有功能的情况下进行修补的能力。关键在于，Mistral Large 4 在处理这些合法的防御性安全任务时不会无故拒绝。
安全性随着能力的提升而增强。我们模型内置的安全护栏能够区分防御性和恶意/进攻性网络安全任务。它在网络基准测试中保持了强劲的性能，同时对恶意网络请求的平均拒绝率达到 96.6%，这是我们测量的开源模型中最高的。
在预览以来的这段时间里，我们仔细监控了实时流量中与网络安全相关的用法。虽然我们观察到模型拒绝配合针对实时目标的黑盒攻击，但我们看到了进行真实攻击的提升潜力。我们正在相应地修改我们的安全护栏，并在未来几个月内持续更新它们。
我们还向企业设计合作伙伴（如 Orange Cyberdefense）提供专门针对高级网络安全工作流优化的 Mistral Large 4 变体的预览访问权限，包括渗透测试、漏洞评估和红队演练。访问需经过严格的审查和合同保障，测试仅限于受控环境中的自有或明确授权的系统。
**组织部署的智能体必须是安全的。**无论是部署给内部成员还是外部用户，智能体都不应让您面临系统提示词泄露或恶意代码注入等威胁。RL 增加了我们在 B3 Agent Security Benchmark 上的收益，Mistral Large 4 抵御了 97.1% 的攻击，而预览版为 93.3%。这使其处于 OSS 前沿的顶部，意味着任何托管该模型的人都可以运行安全的智能体而无需额外的调整。
Mistral AI 的这一重大投资巩固了 La Banque Postale 一年多前做出的选择，并强调了我们要共同建立的长期战略伙伴关系的相关性：支持出现一个开放且欧洲主权 AI 领导者。作为一家面向公众和公民的银行，技术主权并非可有可无；它是维持客户信任的关键先决条件。这一开放模型的引入解锁了新的机会和使用场景。通过与 Mistral AI 合作，我们的雄心是将人类专业知识和数字技术的最佳结合，长期为客户提供更优质的体验。
借助 Mistral Large 4，这种控制权始于模型本身。组织有能力在其自己的基础设施上检查、定制和部署该模型。他们可以使用并改进 Mistral Large 4，而无需依赖 Mistral 通过 API 提供服务。Mistral Large 4 在欧洲从头开发和训练，使组织能够清楚了解他们所构建技术的起源。
主权并不总是需要自托管；有时它只需要满足您控制要求的部署。对于偏好托管服务的组织，Mistral Large 4 可通过 Microsoft Foundry 获得，Azure 开发人员可选择私有部署选项。
Microsoft 和 Mistral AI 都致力于扩大下一代 AI 的客户选择。作为首家将 Mistral Large 4 带给客户的云提供商，Microsoft 正通过 Microsoft Foundry 提供 Mistral 最新的开源权重模型，这是一个专为大规模评估、部署和操作模型而构建的企业 AI 平台。对于推进主权 AI 战略的组织，这种组合为其数据、基础设施和 AI 操作提供了更大的灵活性和控制权。
查看原文
Thinking
Summary
Mistral released Mistral Large 4 (Le Chonk), a 1-trillion-parameter multimodal model, now available with improved coding, cybersecurity, and reasoning efficiency. It leads open-weight benchmarks in coding and defensive security while offering three reasoning modes for cost-performance balance. Fully open-weight and European-trained, it empowers sovereign AI with customizable, self-hosted deployment options.
We release the weights of Le Chonk, advancing the frontier of open models in agentic coding and cybersecurity.
Earlier this month, we released Mistral Large 4 (Le Chonk), our flagship 1-trillion-parameter, natively multimodal model, in public preview. The response to the preview has exceeded our expectations, and we're grateful to our developer community, early-access customers, and partners for their support, collaboration, and feedback.
Since then, we've put that feedback to work while continuing to scale reinforcement learning, driving massive improvements in coding and defensive cybersecurity. Gains continue to accelerate, with no signs of saturation.
Today, we're releasing the weights of this improved Mistral Large 4. The new model also features substantial gains in reasoning efficiency that give developers greater performance while lowering inference costs.
Mistral Large 4 ranks as the best open-weight model for coding and among the world's most capable open models overall. Trained from scratch on European infrastructure with clear lineage from inception to release, it gives organizations a trusted foundation for Sovereign AI
Mistral is defining the sovereign AI category and leading the open frontier. We're committed to building the world's most capable models and releasing them openly, giving developers, enterprises, and institutions the power to own, customize, and control their AI.
Built for the Builders.
Mistral Large 4 entered public preview with strong results across leading coding benchmarks. Just three weeks of additional reinforcement learning have since propelled it to the forefront of open-weight coding, taking the lead on FrontierCode and Terminal-Bench while matching the best open-weight models on the Artificial Analysis Coding Agent Index with substantially greater token efficiency.
On Cognition's private FrontierCode 2.0 benchmark, which assesses end-to-end code quality and whether a repository maintainer would merge the agent's change, Mistral Large 4 leads with a score of 19.7%.
Cognition - FrontierCode 2.0
Pushing the pareto frontier for 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, 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.
Artificial Analysis Coding Agent Index vs. Token Efficiency
Coding Agent Index scores and token efficiencies for reasoning levels available on Artificial Analysis
The gains extend to terminal-based coding workflows. On Terminal-Bench 4.0, Mistral Large 4 improved by 13 percentage points to 41% compared to its public preview, taking the lead among open-weight models while reducing output tokens by 30%.
Artificial Analysis Coding Agent Index - Terminal-Bench 4.0
To give developers greater control over this tradeoff between performance and cost, Mistral Large 4 now supports three reasoning modes:
Mode
Recommended Use
Low
Simple tasks; replaces ‘None’ reasoning mode
Medium
Everyday tasks
High (default)
Hardest use cases requiring high precision and thorough reasoning
Beyond benchmarks, Mistral Large 4's coding capabilities are validated by human preference. In Surge Code's blind pairwise evaluations, Mistral Large 4 was preferred over leading open-weight models including GLM-5.3, Kimi K3, and DeepSeek V4.1 Flash, ranking behind only Claude Opus 5.5 among the models evaluated.
Surge Code - Human preference on code
Blind side-by-side win rates against other models
-
Mistral Large 4 win
-
Other model
Share of prompts where graders preferred Mistral Large 4's answer over the opponent's.
Built for the Defenders.
Since its preview, we've been working with leading cybersecurity vendors to evaluate Mistral Large 4 in real-world environments to test Mistral Large 4’s strengths in vulnerability reproduction and patching.
Mistral Large 4 scores 51% on the Artificial Analysis Cyber Index, which measures a model’s ability to discover vulnerabilities in a codebase, reproduce and validate them, and patch them without breaking existing functionality. Crucially, Mistral Large 4 handles these legitimate defensive security tasks without unnecessary refusals.
Artificial Analysis - Cyber Index
Safety has improved alongside capability. Our model’s built-in guardrails distinguishes between defensive and malicious, offensive cybersecurity tasks. It retains strong performance on Cyber benchmarks, while achieving an average refusal rate of 96.6% on malicious cyber requests, the highest we measure among open models.
Refusal of harmful cyber requests
Average refusal on the cyber-related prompts of JailbreakBench, StrongREJECT and AgentHarm
Here’s what global leaders in cybersecurity have to say about Mistral Large 4:
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.
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, including penetration testing, vulnerability assessment, and red-teaming. Access is subject to rigorous vetting and contractual safeguards, with testing restricted to owned or explicitly authorized systems in controlled environments.
Meanwhile, customers are tailoring Mistral Large 4 with proprietary security data, policies, detection logic, and threat environments using Mistral Forge. For well-defined cybersecurity tasks, it can be further optimized into smaller, task-specific variants for greater accuracy, making security applications more effective.
Learn more about sovereign AI for cybersecurity and how to work with us.
Built for Work.
Mistral Large 4 has continued improving on a range of broad enterprise-use metrics.
For finance-related work, we improved the model’s ability to perform real-world workflows, especially ones involving the creation and editing of spreadsheets for finance and accounting use cases. This is reflected in our Finch scores climbing to 73.3%, second only to Kimi K3 for open models.
FinWorkBench (Finch)
Web browsing is a critical part of a majority of knowledge work. Our RL pipeline enabled us to reach 84.5% on BrowseComp (starting from an SFT checkpoint that was achieving 75%). This brings Mistral Large 4 at a higher performance than models like GLM-5.3 or DeepSeek Flash 4.1 1017.
Web browsing - BrowseComp
Agents deployed by an organization need to be secure. Whether deployed to internal members or external users, agents shouldn’t open you up to threats like system prompt exfiltration or malicious code injection. RL increased our gains on the B3 Agent Security Benchmark, with Mistral Large 4 resisting 97.1% of attacks vs 93.3% in Preview. This places it at the top of the OSS frontier, meaning anyone hosting the model can run a safe agent without additional tuning.
B3 Agent security benchmark - Attack resistance
This major investment by Mistral AI reinforces the choice La Banque Postale made over a year ago and underscores the relevance of the long-term strategic partnership we are building together: supporting the emergence of an open and sovereign European AI leader. For a public and citizen bank, technological sovereignty is not optional; it is a key prerequisite for maintaining our customers' trust. The introduction of this open model unlocks new opportunities and use cases. By partnering with Mistral AI, our ambition is to combine the best of human expertise and digital technology to deliver an enhanced experience of both customers and employees over the long term.
Zakaria Moursli, Executive Vice President, Operational Performance, Payments and Digital Services, La Banque Postale
See it in Action
Cyber Vulnerability Triage: an agent reads incoming vulnerability reports, investigates the source code and product context, reproduces the flaws in an isolated sandbox, then writes and iterates on a patch until a canary exploit no longer works.
Agentic webdev: with no human help, Mistral Large 4 builds Orbital Trace, a live 3D map of roughly 19,000 satellites and debris pieces orbiting Earth. It pulls real tracking data, writes the app, uses it in a browser to catch its own bugs and slowdowns, then fixes what an independent reviewer flags until all requirements pass.
Quantum chemistry solver: Mistral Large 4 writes, from scratch, software that calculates where electrons sit in a molecule, and checks its answers against a standard chemistry tool. It then builds a 3D page around it: pick a molecule such as benzene or water, drag a slider to pull it apart, and watch the electrons regroup as the bond breaks.
Built for Sovereignty.
As organizations build AI around their knowledge and workflows, sovereignty means that they retain control of the models they depend on and the intelligence they create.
With Mistral Large 4, that control begins with the model itself. Organizations have the ability to inspect, customize, and deploy the model on their own infrastructure. They can use and improve Mistral Large 4 without depending on Mistral to serve it through an API. Mistral Large 4 was developed and trained from scratch in Europe, giving organizations clarity about the origins of the technology they’re building on.
Control also extends to how AI operates inside an organization. Organizations can customize Mistral Large 4 around their proprietary data with Mistral Forge, and they can build, govern, and improve the resulting agents and applications using Mistral Studio. Mistral’s Applied AI experts can work alongside customer teams to build and deploy these tailored systems. Through that work, teams understand the architecture and decisions behind what they’ve built, along with the methods used to train and operate it, so they can run and improve their AI independently.
Sovereignty does not always require self-hosting; sometimes it just requires deployments that meet your control requirements. For organizations that prefer a managed service, Mistral Large 4 is available through Microsoft Foundry, with private deployment options for Azure developers.
Microsoft and Mistral AI share a commitment to expanding customer choice in next-generation AI. As the first cloud provider to bring Mistral Large 4 to customers, Microsoft is making Mistral’s latest open-weight model available through Microsoft Foundry, an enterprise AI platform built for evaluating, deploying, and operating models at scale. For organizations advancing sovereign AI strategies, this combination offers greater flexibility and control across their data, infrastructure, and AI operations.
Tina Schuchman, Corporate Vice President, Foundry, Microsoft
Start building with Le Chonk
Mistral Large 4 is available to try, deploy, and customize today.
Get started in Mistral Studio, through the Mistral API, or on OpenRouter. For full control, download the open weights from Hugging Face and deploy on your own infrastructure with vLLM or SGLang. The model card and license have everything you need to get started — and for those who want to go deeper, a full technical report is coming soon.
You can also try the new model in Vibe Code; give it a feature to build or a bug to fix in your own repository.
Mistral Large 4, is also available through NVIDIA-hosted endpoints, as an optimized NVIDIA NIM container for serving with vLLM. Developers can use the open source NeMo AutoModel library for customization or the open source TensorRT LLM library for high-throughput serving.
Enterprises can also tailor Mistral Large 4 to their proprietary data, use cases, and guardrails with Mistral Forge. Talk to our team to learn more.
The model is available to use under the Mistral Ecosystem Open-Weight license, a modified MIT license. 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.
Starting today, Mistral Large 4 is priced at $0.70 input / M, $0.07 cached input / M, and $2.10 output / M tokens.
We encourage the community to put Mistral Large 4 to the test. Run your own evaluations, push its limits, and share your findings with us on X, Discord, or GitHub.
What's Next
Mistral Large 4 will serve as the foundation for a new generation of specialized and optimized Mistral models. Building on recent releases such as Voxtral and Timestral, we'll continue expanding the frontier of open-weight AI, continuing to give developers and enterprises greater freedom to build, customize, and own their intelligence.
Studio