Unverified90% confidenceFactTime unknown
Sonnet并非Anthropic的顶级模型，Opus才是Anthropic的顶级模型
1
Sources
90%
Confidence
Long-term
Relevance
6/2/2026
First Seen
Sources
Claude Sonnet 4.5 vs GPT-5 Codex：软体物理编程实战对决
bilibiliAIDeepCoder
Related Entities
Related Claims
UnverifiedAnthropic内部有一个代号为Mythos的更强模型但选择暂不发布64% similarUnverifiedAnthropic的模型命名体系中，Opus、Sonnet、Haiku分别对应高、中、低三个能力与成本梯度54% similarUnverifiedAnthropic缺乏足够的算力来慷慨地服务其最强的模型（Fable）54% similarUnverifiedAnthropic的基础模型Methos是Fable 5的基础，Anthropic曾声称Methos'太危险了，不能交给大众'54% similarUnverified阿里方面称该通义千问模型综合表现仅次于Anthropic53% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/37931API
curl https://kongchang.com/api/v1/knowledge/claims/37931MCP
get_claim(id=37931)