読み込み中...
AI パワーランキングAI パワーランキングAI パワーランキング読み込み中...
Anthropic's Claude Opus 4 achieves 72.5% on SWE-bench, setting a new state-of-the-art record for autonomous software engineering capabilities.
72.5%
SWE-bench Performance
Note: This analysis was compiled by AI Power Rankings based on publicly available information. Metrics and insights are extracted to provide quantitative context for tracking AI tool developments.