Unverified50% confidenceBenchmarkExact time
在Agent's Last Exam基准测试中，GPT-5.6 Soul（extra high档）得分近54%，Fable（max档）为45%
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/12/2026
First Seen
Valid until: 10/10/2026
Sources
GPT-5.6深度评测：性价比碾压竞品，但越狱安全漏洞不容忽视
bilibiliTOP高效学习法7/10/2026
Related Claims
Unverified在Agent Test: Last Exam测试中，Soul得分53.6分，Anthropic的Opus 4.5得分40.5分82% similarUnverifiedGPT-5.6在Agent's Last Exam测试中以53.6分创下历史新高，领先Claude Opus 4.5达13.1分79% 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77% similarUnverified在Health Bench Professional上，Soul达到60.5分，比GPT-5.5高出8.7分72% similarPartially VerifiedClaude Opus 4.8在SWE-bench基准测试中得分69.2%，GPT 5.5为58.6%，Google Gemini为54.2%70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/491033API
curl https://kongchang.com/api/v1/knowledge/claims/491033MCP
get_claim(id=491033)