SWE-Milestone — leaderboard
Metric: Milestone Score (%). Source: swe-milestone.com. 23 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GLM-5.2 | 45.92 |
| 2 | Claude Opus 4.7 (xHigh) | 41.29 |
| 3 | GPT-5.5 (xHigh) | 38.85 |
| 4 | Claude Opus 4.6 | 38.57 |
| 5 | Kimi K2.6 | 36.8 |
| 6 | GPT-5.4 (xHigh) | 34.53 |
| 7 | Gemini 3.5 Flash | 33.13 |
| 8 | Claude Sonnet 4.6 | 30.66 |
| 9 | GPT-5.3 Codex (xHigh) | 30.53 |
| 10 | GLM-5.1 | 29.25 |
| 11 | Claude Opus 4.5 | 26.43 |
| 12 | DeepSeek V4 Pro | 25.26 |
| 13 | Gemini 3.1 Pro (Preview) | 25.25 |
| 14 | Gemini 3 Flash | 24.82 |
| 15 | GPT-5.2 (xHigh) | 24.62 |
Interactive version: aibenchmarks.dev/benchmark?slug=swe-milestone · How the rankings work · Data refreshed daily, snapshot 2026-07-20.