Mistral Large 4 solves 18 of 19 challenges in cybersecurity CTF speedrun

Mistral AI demonstrated Mistral Large 4 solving 18 out of 19 challenges in a Capture the Flag (CTF) speedrun evaluation designed to test reasoning across complex cybersecurity tasks.

According to Mistral AI, the test evaluated the model's capabilities using tool calls, with solve times and executions measured from actual runs.