Unverified50% confidenceBenchmarkExact time
在SWE-bench Full版本上，多数模型的通过率仍低于20%
1
Sources
50%
Confidence
Long-term
Relevance
7/11/2026
First Seen
Sources
GPT-5.6三款模型接入Cursor：Sol拿下67.2%基准分
twittercursor_ai7/9/2026
Related Claims
VerifiedIn early 2024, the strongest AI models scored only in the 10%–20% range on SWE-Bench.75% similarUnverifiedWhen SWE-bench was first released, the pass rate of top models was below 2%; the pass rate of leading systems has now surpassed 50%.74% similarUnverifiedThe earliest models scored less than 20% on the GDPVal benchmark, far below human performance.72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/483582API
curl https://kongchang.com/api/v1/knowledge/claims/483582MCP
get_claim(id=483582)