Unverified60% confidenceFactExact time
在AID基准测试中，o1 high（对应o1 pro）的编程能力得分超过60，优于o3-mini-high，但两者非常接近
2
Sources
60%
Confidence
Long-term
Relevance
2/5/2025
First Seen
Sources
o1、o1 pro与o3-mini-high编程能力深度对比：Deep Research实测分析
bilibilikate人不错2/5/2025
Related Entities
Related Claims
Unverified在Pass@4通过率测试中，o1 pro的表现显著优于o176% similarUnverified在SWE-Bench评分中，o1的得分远高于o3-mini medium，但略低于o3-mini-high74% similarUnverified在Codeforces评分中，o1的得分远低于o3-mini-high73% similarUnverifiedCodex One在SWE-bench上的表现略优于O3 High，在OpenAI内部软件工程任务上准确率达到75%，而O3 High为70%66% similarUnverified在编程基准测试中，Codex-1的表现超越了Claude 3.7和O3 High64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/655API
curl https://kongchang.com/api/v1/knowledge/claims/655MCP
get_claim(id=655)