Unverified50% confidenceBenchmarkExact time
Soul标准模式得分88.8%，开启Ultra后跃升至91.9%，比Max模式多涨2.7个百分点
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/8/2026
First Seen
Valid until: 10/6/2026
Sources
GPT-5.6编程基准91.9%：三档模型选择与Ultra模式避坑指南
bilibili费曼学AI6/28/2026
Related Claims
Unverified旗舰Soul在Terminal Bench 2.1基准上取得91.9%的成绩70% similarUnverified在Health Bench Professional上，Soul达到60.5分，比GPT-5.5高出8.7分68% similarUnverified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%61% similarUnverified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%61% similarUnverified在前端与视觉任务上Soul只能拿到六到七分，而Fable五能拿到九到十分61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/230625API
curl https://kongchang.com/api/v1/knowledge/claims/230625MCP
get_claim(id=230625)