Unverified60% confidenceFactTime unknown
Early GPT-4 achieved a resolution rate of less than 2% on SWE-bench, while recent top models have surpassed 50%.
2
Sources
60%
Confidence
Long-term
Relevance
5/31/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedGPT-4o is priced approximately 50% lower than GPT-4.79% similarUnverifiedGPT-5.6在SWE-bench评测中以不到竞品一半的成本实现更优效果和更高解决率79% similarUnverifiedEven the most advanced current models such as GLM-5.2, Claude-Sonnet-4.6, and GPT achieve only just over half marks on this benchmark.78% similarUnverifiedGPT-5.4 was reversed by a "trusted source" on 44.7% of questions, Grok-4.20 had a reversal rate as high as 87.5%, while Gemini-3.1-Pro was nearly unaffected at just 0.6%.78% similarUnverifiedGPT-4 发布时在 MATH 数据集上的正确率为 42.5%76% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/14582API
curl https://kongchang.com/api/v1/knowledge/claims/14582MCP
get_claim(id=14582)