Mistral's open-source Leanstral 1.5 aces formal math benchmarks and catches real bugs in code

Mistral AI has released Leanstral 1.5, an open-source language model specifically trained to assist with formal verification in the Lean 4 programming language. Beyond its performance on mathematical benchmarks, the model successfully identified five previously unknown bugs while auditing 57 open-source code repositories. This development marks a shift toward using large language models for rigorous automated software testing and proof engineering, potentially improving code reliability by automating the detection of errors that traditional debugging methods might overlook.