Unverified50% confidenceBenchmarkExact time
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%
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/13/2026
First Seen
Valid until: 10/11/2026
Sources
GPT-5.6三款新模型发布：Luna、Terra、Sol主打智能体长任务
rss7/9/2026
Related Claims
Unverified在Artificial Analysis的Omniscience幻觉指标上，GPT-5.6 Sol Max高达89%，高于Claude Fable的55%和GLM 5.2的28%81% similarUnverified在Agents' Last Exam评测中，GPT-5.6 Sol取得53.6分，比采用自适应推理的Claude Fable 5高出13.1分81% similarUnverified在Terminal Bench 2.1基准测试上，GPT-5.6 Sol的Ultra模式接近92%，超越竞品的88%表现79% similarUnverifiedFable 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78% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/497876API
curl https://kongchang.com/api/v1/knowledge/claims/497876MCP
get_claim(id=497876)