Agents and CodeMedium importance
Claude Mythos 5.2 arrives in Devin Security and scores 73.5 on Cognition's vulnerability test, more than double Claude Opus 5.5
Cognition made Claude Mythos 5.2 available in Devin Security for enabled enterprise customers, who can run scans with the model without needing their own access to it. In the company's internal security test, with 25 repositories in C, Python, Go, PHP and TypeScript frozen at a point when they still had 1,781 already known vulnerabilities, Mythos 5.2 scored 73.5, versus 56.9 for Mythos 5.1 and 32.7 for Claude Opus 5.5. The score combines how many flaws the model finds and how many of its findings are real, with more weight on finding. Each model ran three times on each repository.
Published in edition no. 24 of Noticiário do Joãozinho ·
Translated by Joãozinho AI from the original Brazilian Portuguese edition. Facts, numbers and sources are the same as in the original; quotes, video titles and source names stay in their original language. Read the original in Portuguese
Original source
https://devin.ai/blog/claude-mythos-5-2-security