23 Daily — French AI startup Mistral released its trillion-parameter open-source AI model Mistral Large 4 (ML4) on October 29, 2026, publicly sharing the model weights to enable businesses and governments to download and deploy the system independently. According to techcrunch.com, this release highlights intensifying competition between open-weight AI developers like Mistral and closed-model providers such as OpenAI and Anthropic, with particular relevance in cybersecurity.
Mistral offers a controllable and cost-efficient AI cybersecurity solution
Nicknamed “Le Chonk,” ML4 aims to provide advanced AI capabilities while maintaining full control for organizations over their data and systems. Mistral CEO Arthur Mensch stated AI should be "a tool for empowerment, not dominance." The company claims ML4 achieves “state-of-the-art performance” at roughly half the computational and capital cost of closed labs.
ML4 also outperforms all other open models on Artificial Analysis' Coding Agent Index, particularly in coding-related tasks. Performance-wise, it is comparable to the well-regarded Chinese open model GLM-5.3 but uses 47% fewer tokens per task, making it significantly more economical.
Open weights empower organizations to deploy AI defenses autonomously
Mistral provides ML4 weights and API access via Hugging Face, facilitating internal adoption by organizations. Mistral’s VP of Science, Pierre Stock, explained that having access to the weights means businesses and governments do not rely on external vendors and can use advanced AI to defend against cyberattacks.
He added that ML4 can identify, coordinate, and respond to major security incidents in real-time within a company's own infrastructure. Stock emphasized the importance of widespread access to AI capabilities to strengthen the security ecosystem and cautioned against fearmongering regarding potential malicious AI use.
Editor's take: Opportunities and validation for open-source AI in cybersecurity
Mistral’s decision to open-source ML4 weights provides enterprises and governments greater ownership of AI-powered cybersecurity defenses, reducing reliance on third-party providers and enhancing privacy and operational flexibility. This represents significant autonomy in cyber defense systems.
Entrepreneurs should watch this trend to develop customized solutions tailored to industry or security-specific needs. To validate benefits, organizations can trial ML4 in real-world monitoring and intrusion detection scenarios, comparing its performance and responsiveness to closed models to assess cost savings and control improvements.