Mistral AI Publicly Releases Large 4, Targets Cybersecurity Users
The 1 trillion-parameter model now includes stronger coding and cyber-defense results after three extra weeks of reinforcement learning, Mistral said.
- On Thursday, French AI startup Mistral released its Large 4 model publicly, designed to help organizations identify and fix cyber-security vulnerabilities. The trillion-parameter system activates 52 billion parameters per task.
- Previously nicknamed 'Le Chonk,' the model underwent three weeks of reinforcement learning guided by feedback from governments and enterprises. Mistral says it is the 'world's best' open-weight model for automated coding and cybersecurity.
- ML4 scored 79% on CyberGym-E2E and 41% on Terminal-Bench 4.0, while using 47% fewer tokens per task than competitors. The model was trained on 3,800 Nvidia Grace Blackwell GPUs in Europe.
- Mistral Vice President of Science Pierre Stock framed the release as ensuring continuity. 'AI is going to be like electricity,' Stock said, arguing companies should not depend on providers who can cut off access.
- Experts warn of dual-use risks as the model offers defensive capabilities. Sophos senior director of threat intelligence Rafe Pilling noted freely available models 'can also be abused as well as used for productive purposes,' requiring rigorous monitoring.
17 Articles
17 Articles
Mistral AI releases Large 4, a 1-trillion-parameter open-weight model with a cybersecurity focus - Tech Startups
French AI startup Mistral AI publicly released its 1-trillion-parameter Large 4 model on Thursday, letting companies and developers run advanced AI on their own infrastructure without relying on cloud services controlled by U.S. technology giants. The release puts cybersecurity at […] The post Mistral AI releases Large 4, a 1-trillion-parameter open-weight model with a cybersecurity focus first appeared on Tech Startups.
With Mistral Large 4 the French start-up releases the "best open model" in the world, at least for programming and cybersecurity. This is good news for Mistral, but also for Europe.
Channel News Asia
Mistral AI publicly releases Large 4, targets cybersecurity users
Oct 29 : French AI startup Mistral released its Large 4 model to the public on Thursday, saying it was designed to help organisations identify and fix cyber-security vulnerabilities.Unlike leading AI systems such as Google's Gemini, which are mainly accessed through online services, the model can be downloade
Coverage Details
Bias Distribution
- 56% of the sources are Center
Factuality
To view factuality data please Upgrade to Premium
