# SWE-Bench Pro — AI model rankings

Can the AI fix real bugs in real software? It's handed actual problems from open-source projects and has to write code that genuinely solves them. Higher is better.

18 tracked models have a published SWE-Bench Pro score. Higher is better. Scores are as published at each model's release.

## Ranking

| Rank | Model | Developer | Score | Released |
| --- | --- | --- | --- | --- |
| 1 | Claude Fable 5 | Anthropic | 80.3% | Jun 9 2026 |
| 2 | Claude Opus 4.8 | Anthropic | 69.2% | May 28 2026 |
| 3 | Qwen3.8-Max | Qwen | 67.7% | Aug 3 2026 |
| 4 | Grok 4.5 | SpaceXAI | 64.7% | Jul 8 2026 |
| 5 | Claude Opus 4.7 | Anthropic | 64.3% | Apr 16 2026 |
| 6 | Claude Sonnet 5 | Anthropic | 63.2% | Jun 30 2026 |
| 7 | GLM-5.2 | Z.ai | 62.1% | Jun 16 2026 |
| 8 | Muse Spark 1.1 | Meta | 61.5% | Jul 9 2026 |
| 9 | Qwen3.7-Max | Qwen | 60.6% | May 20 2026 |
| 10 | GPT-5.5 | OpenAI | 58.6% | Apr 23 2026 |
| 11 | Gemini 3.5 Flash | Google | 55.1% | May 19 2026 |
| 12 | Muse Spark | Meta | 55% | Apr 8 2026 |
| 13 | Gemini 3.1 Pro | Google | 54.2% | Feb 19 2026 |
| 13 | Gemini 3.5 Flash-Lite | Google | 54.2% | Jul 21 2026 |
| 15 | Composer 2.5 | Cursor | 54% | May 18 2026 |
| 16 | Gemini 3.0 Flash | Google | 49.6% | Dec 17 2025 |
| 17 | Qwen3.6 | Qwen | 49.5% | Apr 16 2026 |
| 18 | Qwen3-Coder-Next | Qwen | 44.3% | Feb 3 2026 |


---

Canonical page: https://aireleasetracker.com/benchmark/swe-bench-pro
Full dataset: https://aireleasetracker.com/llms-full.txt · JSON: https://aireleasetracker.com/models.json
Source: AI Release Tracker (https://aireleasetracker.com). Benchmark scores are the figures published by the releasing lab at launch.