Artificial Analysis 榜单以 FrontierCode v1.1 替代 DeepSWE v1.1

AI 导读

Artificial Analysis 的 Index 榜单以 Cognition 的 FrontierCode v1.1 替代 DeepSWE v1.1。该基准包含来自真实开源仓库的 100 个软件工程任务，并采用私有测试集，以维护者是否会合并改动作为评分标准。

正文 · 原文

FrontierCode v1.1 replaces DeepSWE v1.1 in the Index. Cognition's benchmark has 100 software engineering tasks on real open-source repositories with a private test set, graded on whether a maintainer would merge the change.

Claude Code with Claude Opus 5.5 (max) scores 56.0%, ahead of Codex with DeepSeek V4.1 Flash (max) at 52.6%, Codex with GPT-6.1 Sol (max) at 49.4% and Codex with GPT-6 Astra (max) at 49.3%. Below the top, scores are tightly packed: 15 of the 24 configurations on the Index score between 43% and 49%.