OpenCodePapers

agents-on-frontierswe

Agents
Dataset Link
Results over time
Click legend items to toggle metrics. Hover points for model names.
Leaderboard
PaperCodeDominanceAverage RankModelNameReleaseDate
FrontierSWE Leaderboard✓ Link89.592.35Claude Fable 5 (Claude Code)2026-06-09
FrontierSWE Leaderboard75.114.24Claude Opus 4.8 (Claude Code)2026-05-28
FrontierSWE Leaderboard✓ Link74.434.32GLM-5.2 (Claude Code)2026-06-16
FrontierSWE Leaderboard72.624.56GPT-5.5 (Codex)2026-04-23
FrontierSWE Leaderboard63.355.76Claude Opus 4.7 (Claude Code)2026-04-16
FrontierSWE Leaderboard56.116.71Claude Opus 4.6 (Claude Code)2026-02-05
FrontierSWE Leaderboard54.306.94GPT-5.4 (Codex)2026-03-05
Composer 2 Technical Report40.278.76Composer 2.5 (Cursor CLI)2026-03-25
FrontierSWE Leaderboard39.598.85Gemini 3.1 Pro (Gemini CLI)2026-02-19
FrontierSWE Leaderboard✓ Link30.5410.03GLM-5.1 (Claude Code)2026-04-03
FrontierSWE Leaderboard✓ Link28.9610.24DeepSeek V4 Pro (Claude Code)2026-04-26
FrontierSWE Leaderboard✓ Link26.7010.53Kimi K2.6 (Kimi CLI)2026-04-20
FrontierSWE Leaderboard✓ Link26.0210.62Kimi K2.5 (Kimi CLI)2026-02-02
FrontierSWE Leaderboard✓ Link22.4011.09Qwen3.6-Plus (Qwen Code)2026-04-21