AIFortuneabout 1 hour ago
Mistral Large 4 Tops Open AI Coding Benchmarks
4 min read
French AI lab Mistral released Large 4, a 1 trillion-parameter model, on Thursday via Hugging Face. It ties with China's GLM-5.3 on Terminal-Bench 4.0, jumping from 28% to 41% accuracy after reinforcement learning. Mistral claims it uses 47% fewer tokens than rivals while remaining the best open model for coding and cyber defense.
Level
Hype check
Tap to vote and see what everyone thinks.
In this storyMistral
