{"generated_at":"2026-09-12T05:44:58.492Z","source":{"label":"SWE-bench leaderboard","url":"https://www.swebench.com/","retrieved_at":"2026-09-12T05:44:33.617Z","board_last_updated":"2026-02-26","board_stale_days":198,"freshness_note":"We re-fetch this board daily, but SWE-bench itself has not accepted a new Verified submission since the date in board_last_updated. Retrieval date is not data freshness."},"boards":{"Multilingual":{"count":13,"top":[{"name":"Gemini 3 Flash","resolved_pct":72.7,"date":"2026-02-13","checked":true},{"name":"Claude 4.6 Opus","resolved_pct":72,"date":"2026-02-13","checked":true},{"name":"Claude 4.5 Opus","resolved_pct":70.7,"date":"2026-02-13","checked":true},{"name":"GLM 5","resolved_pct":69.7,"date":"2026-02-13","checked":true},{"name":"Gemini 3 Pro","resolved_pct":68.7,"date":"2026-02-13","checked":true},{"name":"MiniMax M2.5","resolved_pct":68.3,"date":"2026-02-16","checked":true},{"name":"Kimi K2.5","resolved_pct":67.3,"date":"2026-02-13","checked":true},{"name":"Claude 4.5 Sonnet","resolved_pct":67,"date":"2026-02-13","checked":true},{"name":"GPT 5.2 (high)","resolved_pct":66.7,"date":"2026-02-13","checked":true},{"name":"GPT 5.2 Codex","resolved_pct":66.3,"date":"2026-02-20"},{"name":"Claude 4.5 Haiku","resolved_pct":64.7,"date":"2026-02-13","checked":true},{"name":"DeepSeek V3.2","resolved_pct":59,"date":"2026-02-13","checked":true},{"name":"GPT 5 mini","resolved_pct":39.7,"date":"2026-02-13","checked":true}],"newest_entry":"2026-02-20","oldest_entry":"2026-02-13","verification":{"maintainer_checked":12,"submitter_reported":1,"note":"Entries marked checked were re-run by the SWE-bench maintainers; the rest are submitter-reported with logs."}},"Test":{"count":24,"top":[{"name":"Sonar Foundation Agent + Claude 4.5 Opus","resolved_pct":52.62,"date":"2025-12-19"},{"name":"Salesforce AI Research SAGE (bash-only)","resolved_pct":44.25,"date":"2025-10-27"},{"name":"Atlassian Rovo Dev (2025-06-05)","resolved_pct":41.98,"date":"2025-06-05"},{"name":"Amazon Q Developer Agent (v20250405-dev)","resolved_pct":37.1,"date":"2025-05-22"},{"name":"SWE-agent 1.0 (Claude 3.7 Sonnet)","resolved_pct":33.83,"date":"2025-02-27","checked":true},{"name":"Amazon Q Developer Agent (v20241202-dev)","resolved_pct":29.99,"date":"2025-01-31"},{"name":"OpenHands + CodeAct v2.1 (claude-3-5-sonnet-20241022)","resolved_pct":29.38,"date":"2024-11-03","checked":true},{"name":"AutoCodeRover-v2.0 (Claude-3.5-Sonnet-20241022)","resolved_pct":24.89,"date":"2024-11-21"},{"name":"Honeycomb","resolved_pct":22.06,"date":"2024-08-20"},{"name":"Amazon Q Developer Agent (v20240719-dev)","resolved_pct":19.75,"date":"2024-07-21"},{"name":"Factory Code Droid","resolved_pct":19.27,"date":"2024-06-17"},{"name":"AutoCodeRover (v20240620) + GPT 4o (2024-05-13)","resolved_pct":18.83,"date":"2024-06-28"},{"name":"SWE-agent + Claude 3.5 Sonnet","resolved_pct":18.13,"date":"2024-06-20","checked":true},{"name":"AppMap Navie + GPT 4o (2024-05-13)","resolved_pct":14.6,"date":"2024-06-15","checked":true},{"name":"Amazon Q Developer Agent (v20240430-dev)","resolved_pct":13.82,"date":"2024-05-09"},{"name":"SWE-agent + GPT 4 (1106)","resolved_pct":12.47,"date":"2024-04-02","checked":true},{"name":"SWE-agent + GPT 4o (2024-05-13)","resolved_pct":11.99,"date":"2024-07-28","checked":true},{"name":"SWE-agent + Claude 3 Opus","resolved_pct":10.51,"date":"2024-04-02","checked":true},{"name":"RAG + Claude 3 Opus","resolved_pct":3.79,"date":"2024-04-02","checked":true},{"name":"RAG + Claude 2","resolved_pct":1.96,"date":"2023-10-10","checked":true}],"newest_entry":"2025-12-19","oldest_entry":"2023-10-10","verification":{"maintainer_checked":13,"submitter_reported":11,"note":"Entries marked checked were re-run by the SWE-bench maintainers; the rest are submitter-reported with logs."}},"Verified":{"count":180,"top":[{"name":"Sonar Foundation Agent + Claude 4.5 Opus","resolved_pct":79.2,"date":"2025-12-05"},{"name":"live-SWE-agent + Claude 4.5 Opus medium (20251101)","resolved_pct":79.2,"date":"2025-12-15"},{"name":"TRAE + Doubao-Seed-Code","resolved_pct":78.8,"date":"2025-09-28"},{"name":"live-SWE-agent + Gemini 3 Pro Preview (2025-11-18)","resolved_pct":77.4,"date":"2025-11-20"},{"name":"EPAM AI/Run Developer Agent v20250719 + Claude 4 Sonnet","resolved_pct":76.8,"date":"2025-08-04"},{"name":"Atlassian Rovo Dev (2025-09-02)","resolved_pct":76.8,"date":"2025-09-02"},{"name":"Claude 4.5 Opus (high)","resolved_pct":76.8,"date":"2026-02-17"},{"name":"ACoder","resolved_pct":76.4,"date":"2025-08-19"},{"name":"Gemini 3 Flash (high)","resolved_pct":75.8,"date":"2026-02-17"},{"name":"MiniMax M2.5 (high)","resolved_pct":75.8,"date":"2026-02-17"},{"name":"Warp","resolved_pct":75.6,"date":"2025-09-01"},{"name":"Claude 4.6 Opus","resolved_pct":75.6,"date":"2026-02-17"},{"name":"TRAE + Claude Sonnet 4 + Opus 4 + Sonnet 3.7 + Gemini 2.5 Pro","resolved_pct":75.2,"date":"2025-06-12"},{"name":"Harness AI","resolved_pct":74.8,"date":"2025-07-31"},{"name":"Sonar Foundation Agent + Claude 4.5 Sonnet","resolved_pct":74.8,"date":"2025-11-03"},{"name":"Lingxi-v1.5_claude-4-sonnet-20250514","resolved_pct":74.6,"date":"2025-07-20"},{"name":"JoyCode + Claude 4 Sonnet + GPT-4.1","resolved_pct":74.6,"date":"2025-09-15"},{"name":"Refact.ai Agent + Claude 4 Sonnet + o4-mini","resolved_pct":74.4,"date":"2025-06-03"},{"name":"Prometheus-v1.2.1 + GPT-5","resolved_pct":74.4,"date":"2025-10-15"},{"name":"Claude 4.5 Opus (20251101) (medium)","resolved_pct":74.4,"date":"2025-11-24","checked":true}],"newest_entry":"2026-02-26","oldest_entry":"2023-10-10","verification":{"maintainer_checked":60,"submitter_reported":120,"note":"Entries marked checked were re-run by the SWE-bench maintainers; the rest are submitter-reported with logs."}},"Lite":{"count":84,"top":[{"name":"ExpeRepair-v1.0 + Claude 4 Sonnet","resolved_pct":60.33,"date":"2025-06-25"},{"name":"Refact.ai Agent","resolved_pct":60,"date":"2025-04-25"},{"name":"KGCompass + Claude 4 Sonnet (20250514)","resolved_pct":58.33,"date":"2025-09-06"},{"name":"SWE-agent + Claude 4 Sonnet","resolved_pct":56.67,"date":"2025-05-26","checked":true},{"name":"Isoform","resolved_pct":55,"date":"2025-01-14"},{"name":"SemAgent_Multi-v1.0","resolved_pct":51.67,"date":"2025-06-25"},{"name":"Isea","resolved_pct":51.33,"date":"2025-09-11"},{"name":"EntroPO + R2E + Qwen3-Coder-30B-A3B-Instruct","resolved_pct":49.67,"date":"2025-09-01"},{"name":"Blackbox AI Agent","resolved_pct":49,"date":"2024-12-20"},{"name":"Codev","resolved_pct":49,"date":"2025-05-28"},{"name":"Gru(2024-12-08)","resolved_pct":48.67,"date":"2024-12-08"},{"name":"Globant Code Fixer Agent","resolved_pct":48.33,"date":"2024-11-27"},{"name":"ExpeRepair-v1.0","resolved_pct":48.33,"date":"2025-06-13"},{"name":"SWE-agent + Claude 3.7 Sonnet","resolved_pct":48,"date":"2025-02-26","checked":true},{"name":"devlo","resolved_pct":47.33,"date":"2024-11-22"},{"name":"DARS Agent","resolved_pct":47,"date":"2025-02-05"},{"name":"KGCompass + Claude 3.5 Sonnet (20241022)","resolved_pct":46,"date":"2025-06-19"},{"name":"EntroPO + R2E + Qwen3-Coder-30B-A3B-Instruct","resolved_pct":45,"date":"2025-09-01"},{"name":"Kodu-v1 + Claude-3.5 Sonnet (20241022)","resolved_pct":44.67,"date":"2024-12-07"},{"name":"CodeFuse-CGM","resolved_pct":44,"date":"2025-03-10","checked":true}],"newest_entry":"2025-09-11","oldest_entry":"2023-10-10","verification":{"maintainer_checked":25,"submitter_reported":59,"note":"Entries marked checked were re-run by the SWE-bench maintainers; the rest are submitter-reported with logs."}},"Multimodal":{"count":22,"top":[{"name":"GUIRepair + o3 (2025-04-16)","resolved_pct":35.98,"date":"2025-07-01","checked":true},{"name":"Codefuse_Pycfuse_SVR","resolved_pct":35.98,"date":"2025-11-17","checked":true},{"name":"Refact.ai Agent","resolved_pct":35.59,"date":"2025-06-11"},{"name":"OpenHands-Versa (Claude-Sonnet 4)","resolved_pct":34.43,"date":"2025-05-28","checked":true},{"name":"GUIRepair + o4-mini (2025-04-16)","resolved_pct":33.85,"date":"2025-05-31","checked":true},{"name":"OpenHands-Versa (Claude-3.7 Sonnet)","resolved_pct":31.33,"date":"2025-05-09","checked":true},{"name":"GUIRepair + GPT 4.1 (2025-04-14)","resolved_pct":31.14,"date":"2025-05-31","checked":true},{"name":"Zencoder (2025-04-01)","resolved_pct":30.56,"date":"2025-04-01","checked":true},{"name":"GUIRepair + GPT 4o (2024-08-06)","resolved_pct":30.37,"date":"2025-05-31","checked":true},{"name":"Globant Code Fixer Agent","resolved_pct":29.59,"date":"2025-03-25","checked":true},{"name":"Zencoder (2025-03-10)","resolved_pct":27.08,"date":"2025-03-11","checked":true},{"name":"Agentless Lite + Claude-3.5 Sonnet","resolved_pct":25.34,"date":"2025-02-26","checked":true},{"name":"SWE-agent Multimodal + GPT 4o (2024-08-06)","resolved_pct":12.19,"date":"2024-10-06","checked":true},{"name":"SWE-agent + Claude Sonnet 3.5","resolved_pct":12.19,"date":"2024-10-06","checked":true},{"name":"SWE-agent JavaScript + Claude Sonnet 3.5","resolved_pct":11.99,"date":"2024-10-06","checked":true},{"name":"SWE-agent + GPT 4o (2024-08-06)","resolved_pct":11.99,"date":"2024-10-06","checked":true},{"name":"SWE-agent Multimodal + Claude 3.5 Sonnet","resolved_pct":11.41,"date":"2024-10-06","checked":true},{"name":"SWE-agent JavaScript + GPT 4o (2024-08-06)","resolved_pct":9.28,"date":"2024-10-06","checked":true},{"name":"Agentless + Claude 3.5 Sonnet","resolved_pct":6.19,"date":"2024-10-06","checked":true},{"name":"RAG + GPT 4o (2024-08-06)","resolved_pct":6,"date":"2024-10-06","checked":true}],"newest_entry":"2025-11-17","oldest_entry":"2024-10-06","verification":{"maintainer_checked":21,"submitter_reported":1,"note":"Entries marked checked were re-run by the SWE-bench maintainers; the rest are submitter-reported with logs."}}}}