Mistral AI has released Mistral Large 4, also known as "Le Chonk," a multimodal, multilingual Mixture-of-Experts (MoE) model. This model accepts text and image inputs and generates text, boasting approximately 1 trillion total parameters with 49 billion active parameters and a 1 million token context window. Mistral Large 4 demonstrates particular strength in cybersecurity, coding agents, and multimodality, scoring highly on a cyber index test for vulnerability reproduction and patching. AI
IMPACT Sets new SOTA on cybersecurity benchmarks and offers a large context window for complex tasks.
RANK_REASON Frontier-lab model release with system card [lever_c_demoted from frontier_release: ic=1 ai=1.0]
- AgentHarm
- Artificial Analysis Cyber Index
- Hugging Face
- JailbreakBench
- Le Chonk
- llama.cpp
- Mistral AI
- Mistral Ecosystem Open-Weight License
- Mistral Large 4
- MIT
- StrongREJECT
- Transformer Reinforcement Learning
- Unsloth
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
