Unverified50% confidenceBenchmarkExact time
在SWE-bench Pro基准上，千问3.8 27B模型的分数超过了Claude Opus
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
8/18/2026
First Seen
Valid until: 11/16/2026
Sources
Related Claims
UnverifiedClaude Opus 4.8 的 Max 模式额度消耗极高，测试者仅问一句'你是什么模型'就消耗了 2% 的月度额度71% similarUnverifiedClaude Opus 4.6 scored 53.4% on the SWE-Bench Pro benchmark.70% similarUnverifiedQwen 3.6 approaches or even surpasses Claude 4.5 Opus on certain metrics in dimensions such as multi-turn agent-enhanced capabilities68% similarUnverifiedClaude Opus 4.7's production task completion on Rakuten SWE Bench is three times that of Opus 4.6.67% similarUnverifiedClaude Opus 4.7 at an intermediate checkpoint achieves a pass rate of 57%, yet its final accuracy is only 39.8%66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/768403API
curl https://kongchang.com/api/v1/knowledge/claims/768403MCP
get_claim(id=768403)