Coding Agent Capacity on SWE-Lancer
56.1ScoreClaude-Opus-4.5
Evaluation results
| Method | Links | |
|---|---|---|
| Claude-Opus-4.52026.06 | 56.1 | |
| Gemini-3-Flash High2026.06 | 51.7 | |
| Seed2.0 Pro2026.06 | 49.4 | |
| GPT-5.2 High2026.06 | 48.9 | |
| Seed2.0 Lite2026.06 | 47.1 | |
| Claude-Sonnet-4.52026.06 | 45.7 | |
| Gemini-3-pro High2026.06 | 44.3 | |
| GPT-5-mini High2026.06 | 43.1 |