Unverified50% confidenceBenchmarkExact time
On the SWE-bench benchmark (2,294 real GitHub issues), Claude 2 scored 1.96%.
1
Sources
50%
Confidence
Long-term
Relevance
10/6/2026
First Seen
Sources
Related Entities
Related Claims
Unverified根据SWE-RPG基准论文数据，当前顶级编程代理在真实GitHub问题上的平均解决率仅为31.5%64% similarUnverifiedGitHub issue #7714 calculated that injecting the same state into the context window consumes 5,764 tokens, while the semantic content itself requires only 3,587 tokens, meaning a 37.8% token overhead per LLM call.64% similarUnverifiedSWE-bench is based on 2,294 GitHub issues from 12 real open-source projects, and the highest pass rate achieved by current systems is only around 50%.61% similarUnverifiedClaude Code's code contributions account for more than 7% of total GitHub commits, and GitHub processes approximately 3 million code commits per day.60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/981295API
curl https://kongchang.com/api/v1/knowledge/claims/981295MCP
get_claim(id=981295)