Unverified50% confidenceFactExact time
Anthropic称Opus 5是迄今最对齐的模型，遵守宪法的程度超过4.8、Sonnet 5和Fable，欺骗行为率最低
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
8/23/2026
First Seen
Valid until: 11/21/2026
Sources
Related Entities
Related Claims
UnverifiedAnthropic表示Sonnet 4.6在提示注入抵抗能力方面的表现与Opus 4.6相当70% similarUnverifiedClaude Haiku 4.5的错误拒绝率低于Sonnet 4.5和Opus 4.1，是Anthropic迄今最安全的模型之一67% similarUnverified根据Anthropic发布的基准测试数据，Sonnet 4.5在编码能力上略优于Claude Opus 4.166% similarUnverified国产模型中编程能力最强的GLM 5.1，实测水平大致处于Claude Opus 4.5和Sonnet 4.6之间65% similarUnverifiedSonnet 4.6与Opus 4.6相比，在大多数基准测试中表现几乎持平，但推理速度更快、调用成本更低65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/791099API
curl https://kongchang.com/api/v1/knowledge/claims/791099MCP
get_claim(id=791099)