Unverified60% confidenceFactTime unknown
Composer的基准测试数据是内部闭源基准，未与Claude、GPT-5或Gemini直接对比，也未出现在LM Arena或SWE-Bench等外部基准上
2
Sources
60%
Confidence
Medium-term (~90 days)
Relevance
6/2/2026
First Seen
Valid until: 8/31/2026(expired)
Sources
Cursor 2.0 深度解析：自研模型Composer与五大核心功能全面升级
bilibili留给时间检验
Related Entities
Related Claims
UnverifiedComposer 2.5在Terminal Bench 2.0、Suite Bench Multilingual和Cursor Bench等基准测试中，某些测试上已超越Opus和GPT 5.570% similarUnverifiedComposer 2.5在SweepBench多语言测试中与Claude Opus 4.7和GPT 5.5属于同一梯队70% similarUnverified根据Cursor内部Benchmark数据，Composer 2.5的能力仅次于OPPO 4.7 Max和GPT 5.5 Extra High66% similarUnverified测试者评价GPT-5.6在前端设计上已具备类似Kimi和Gemini的审美水准66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/36962API
curl https://kongchang.com/api/v1/knowledge/claims/36962MCP
get_claim(id=36962)