{"type":"post","id":"cwe-bench-v2-more-coverage-quality-and-transparency-c19762","slug":"cwe-bench-v2-more-coverage-quality-and-transparency-c19762","title":"CWE-bench v2: More coverage, quality, and transparency","description":"Collinear AI has released CWE-bench v2, a public benchmark of 100 tasks across 91 vulnerability classes for agents that find and fix security weaknesses in real open-source repositories.","retrieval_nugget":"Collinear AI has released CWE-bench v2, a public benchmark of 100 tasks across 91 vulnerability classes for agents that find and fix security weaknesses in real open-source repositories. Its deterministic checks and runnable dataset make model comparisons reproducible, while reported results expose repeated-attempt and cost-quality trade-offs that teams can test themselves.","published_at":"2026-11-03","updated_at":"2026-11-03","record_date":"2026-11-03","date_kind":"published_at","topics":["gemini","claude"],"entities":["Collinear AI","CWE-bench","Harbor","Gemini 4 Argon","Claude Opus 5.5","DeepSeek V4.1 Flash"],"related_patterns":["ai-engineering-evaluations-are-splitting-into-task-specific"],"novelty_label":"new_angle","what_changed":"Unlike the prior code-review and general benchmark coverage, CWE-bench evaluates agents fixing real repository vulnerabilities across 91 CWE categories with reproducible checks.","prior_refs":["https://newruntime.com/posts/reviewbench-an-open-benchmark-for-ai-code-review-c9dcf4/","https://newruntime.com/posts/a-separate-public-benchmark-analysis-evaluated-23-models-6f6beb/","https://newruntime.com/posts/llamaindex-publishes-boo-bench-a-179-page-benchmark-354b9c/"],"source_url":"https://huggingface.co/blog/cwe-bench","source_urls":["https://huggingface.co/blog/cwe-bench"],"schema_version":"newruntime-agent-readable-v0.2","stable_id":"post:cwe-bench-v2-more-coverage-quality-and-transparency-c19762","status":"published","visuals":[],"editor":{"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]},"routes":{"html":"https://newruntime.com/posts/cwe-bench-v2-more-coverage-quality-and-transparency-c19762/","markdown":"https://newruntime.com/posts/cwe-bench-v2-more-coverage-quality-and-transparency-c19762.md","json":"https://newruntime.com/posts/cwe-bench-v2-more-coverage-quality-and-transparency-c19762.json"},"markdown_url":"https://newruntime.com/posts/cwe-bench-v2-more-coverage-quality-and-transparency-c19762.md","next_reads":[{"type":"topic","path":"/topics/gemini/","url":"https://newruntime.com/topics/gemini/","title":"Gemini","reason":"Explore the gemini topic hub.","media_type":"text/html","markdown_url":"https://newruntime.com/topics/gemini.md","json_url":"https://newruntime.com/topics/gemini.json"},{"type":"topic","path":"/topics/claude/","url":"https://newruntime.com/topics/claude/","title":"Claude","reason":"Explore the claude topic hub.","media_type":"text/html","markdown_url":"https://newruntime.com/topics/claude.md","json_url":"https://newruntime.com/topics/claude.json"},{"type":"collection","path":"/posts/","url":"https://newruntime.com/posts/","title":"Field Notes","reason":"Open the latest editorial analysis.","media_type":"text/html","json_url":"https://newruntime.com/posts.json"},{"type":"collection","path":"/sources/","url":"https://newruntime.com/sources/","title":"Sources","reason":"See the original sources this archive cites.","media_type":"text/html"}]}