Expired50% confidenceBenchmarkExact time
GPT-5.6 Sol's score on the ExploitBench cybersecurity benchmark jumped from 40% with GPT-5.5 to 73.5%.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/11/2026
First Seen
Valid until: 10/9/2026(expired)
Sources
GPT-5.6 Hands-On Deep Dive: ChatGPT and Codex Merge, Full Analysis of the Sol Model
bilibilikate人不错7/10/2026
Related Claims
UnverifiedGPT-5.6 on cybersecurity benchmarks: ExploitBench generation 2 at 73.5% (previous generation 47.9%); ExploitGym two-hour-limit peak pass rate increased from 15.1% to 24.9% (reaching 33.7% at 6 hours); CyberSecBench Pro at 71.2% (previous generation 45.8%).81% similarUnverifiedGPT-5.6 Sol achieves 92.2% on the BrowseComp browsing task benchmark.81% similarUnverifiedGPT-5.6's coding capability reached 91.9% on key benchmarks and its safety evaluation CyberGym score reached 85.6%, both surpassing competitor Nexus 5.80% similarVerifiedSOL's benchmark score is nearly double that of the previous generation GPT-5.5, representing a generational leap, but still trails Fable 5 overall.79% similarUnverifiedIn OpenAI's internal benchmarks, GPT-5.6 Sol leads GPT-5.5 by approximately 16 points on recursive self-improvement (model improvement tasks), and shows a 31% improvement over 5.5 on AI infrastructure optimization (chip design workflows).77% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/484711API
curl https://kongchang.com/api/v1/knowledge/claims/484711MCP
get_claim(id=484711)