Unverified50% confidenceFactExact time
Fable 5在SWE Bench Pro编程测试上得分80.3分，而OpenAI的GPT 5.5仅为58.6分
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/2/2026
First Seen
Valid until: 9/30/2026
Sources
Claude Fable 5深度拆解：连续工作9小时的AI，程序员真要被取代了吗
bilibiliMr克劳德Claude6/11/2026
Related Claims
Unverified在Terminal Bench 2.0基准测试中，OpenAI 4.6得分为58%，GPT 5.4得分为75.1%82% similarUnverified在SWE-Bench Pro编程评测上，Claude Fable 5以80%的成绩领先GPT-5.6 Sol的64.6%78% similarUnverifiedTerminal Bench 2.1测试中GPT-5.6 Ultra得分接近92%，Fable 5为88%，考虑误差范围基本持平77% similarUnverifiedGLM 5.2在Terminal Bench测试中得分81分，GLM 5.1得分63.5分，GPT 5.5得分84分，Opus 4.8得分85分73% similarUnverifiedGLM-5在SWE-Bench Verified基准测试中得分77.8%，位于开源第一梯队72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/61384API
curl https://kongchang.com/api/v1/knowledge/claims/61384MCP
get_claim(id=61384)