Cogent says VR-1.1 beats Claude Code and Codex at vulnerability repair

Cogent Security reported that its model produced passing fixes on 57.9% of LiveCVEBench’s 190 tasks, compared with 45.3% each for Claude Code and Codex. The company-run evaluation tested repairs for software vulnerabilities. According to Unite.AI, the competing tools ran Claude Opus 5.5 and GPT-5.6-Cyber, respectively.

• VR-1.1 extends the remediation capabilities of Cogent’s earlier VR-1 model, according to Unite.AI.

Covered by 3 publishers within 2 hours of the first report.

• RuntimeWire: Cogent's VR-1.1 repairs 57.9% of LiveCVEBench tasks in company run

Reporting3

SA SiliconANGLE Exclusive: Cogent Security’s VR-1.1 model outscores Claude Code and Codex on vulnerability repair · Duncan Riley 9:00 AM

UAI Unite.AI Cogent Reports VR-1.1 Passing Repairs on 57.9% of LiveCVEBench Tasks · Miles Okada, AI & Cybersecurity, AI Research Agent 10:03 AM

RTW RuntimeWire Cogent's VR-1.1 repairs 57.9% of LiveCVEBench tasks in company run · Ryan Merket 10:46 AM

Related

THN Yoba attackers use fake ChatGPT and Claude apps to install crypto miners October 29

THN New VOLTA Infostealer Uses Claude, Gemini, and Kimi AI to Rank Victims by Financial Value November 1

CSN Researchers Hacked Cursor, GitHub Copilot, Codex, VS Code, and Claude Desktop Embedded Browsers October 29

CSN Unauthorized OpenAI Codex and Claude Code Relays Disable AI Sandboxes and Redirect Enterprise Data October 29

Topics Anthropic