Unverified70% confidenceFactTime unknown
文章声称Claude Haiku 4.5在多项核心基准测试中击败了GPT-5和Gemini 2.5 Pro
1
Sources
70%
Confidence
Medium-term (~90 days)
Relevance
6/1/2026
First Seen
Valid until: 8/30/2026
Sources
Claude Haiku 4.5评测：三分之一价格实现旗舰级AI性能
bilibili攒钱换房车的福叔
Related Entities
Related Claims
UnverifiedGPT-5 Mini在同类测试中表现远优于Claude 4.5 Haiku81% similarUnverified在SaaS落地页生成测试中，四个模型（Claude Haiku 4.5、Claude Sonnet 4.0、Gemini 2.5 Pro、GPT-5）全部成功完成任务，但Haiku 4.5生成速度最快81% similarUnverifiedClaude 4.5 Haiku在高中级别编程任务上几乎能和GPT-5打平手78% similarVerifiedGPT-5.5在多项能力上全面超越Claude 4.7和Gemini 3.1 Pro等模型75% similarUnverifiedKimi K2 Thinking在多项基准测试中超越了GPT-5和Claude Sonnet 4.575% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/24403API
curl https://kongchang.com/api/v1/knowledge/claims/24403MCP
get_claim(id=24403)