
Mistral's Leanstral 1.5 achieves 90% on PutnamBench with 1/7th of Claude's compute
Mistral AI open-sources Leanstral 1.5, a model that verifies code and mathematical proofs, achieving 90% on PutnamBench.
News, trends, and insights from the AI frontier.
Latest Apache 2.0 license coverage.