FrontierSWE Leaderboard

Benchmark Model Rank Results
agents-on-frontiersweClaude Fable 5 (Claude Code)#1Dominance: 89.59Average Rank: 2.35
agents-on-frontiersweGLM-5.2 (Claude Code)#2Dominance: 74.43Average Rank: 4.32
agents-on-frontiersweGLM-5.1 (Claude Code)#3Dominance: 30.54Average Rank: 10.03
agents-on-frontiersweDeepSeek V4 Pro (Claude Code)#4Dominance: 28.96Average Rank: 10.24
agents-on-frontiersweKimi K2.6 (Kimi CLI)#5Dominance: 26.70Average Rank: 10.53
agents-on-frontiersweKimi K2.5 (Kimi CLI)#6Dominance: 26.02Average Rank: 10.62
agents-on-frontiersweQwen3.6-Plus (Qwen Code)#7Dominance: 22.40Average Rank: 11.09
agents-on-frontiersweClaude Opus 4.6 (Claude Code)–Dominance: 56.11Average Rank: 6.71
agents-on-frontiersweClaude Opus 4.7 (Claude Code)–Dominance: 63.35Average Rank: 5.76
agents-on-frontiersweClaude Opus 4.8 (Claude Code)–Dominance: 75.11Average Rank: 4.24
agents-on-frontiersweGPT-5.4 (Codex)–Dominance: 54.30Average Rank: 6.94
agents-on-frontiersweGPT-5.5 (Codex)–Dominance: 72.62Average Rank: 4.56
agents-on-frontiersweGemini 3.1 Pro (Gemini CLI)–Dominance: 39.59Average Rank: 8.85