Aider Polyglot
Code editing across Python/JS/Go/Rust/etc. via Aider's benchmark harness.
Top 25 models
| Rank | Model | Score | Captured |
|---|---|---|---|
| 1 | GPT-5 | 88.0% | 2025-08-23 |
| 2 | o3 | 84.9% | 2025-06-28 |
| 3 | Gemini 2.5 Pro | 83.1% | 2025-06-06 |
| 4 | Grok 4 | 79.6% | 2025-07-11 |
| 5 | DeepSeek V3 | 74.2% | 2025-10-03 |
| 6 | o4-mini | 72.0% | 2025-04-16 |
| 7 | Claude Opus 4 | 72.0% | 2025-05-25 |
| 8 | DeepSeek R1 | 71.4% | 2025-06-06 |
| 9 | Claude 3.7 Sonnet | 64.9% | 2025-02-24 |
| 10 | o1 | 61.7% | 2024-12-21 |
| 11 | Claude Sonnet 4 | 61.3% | 2025-05-24 |
| 12 | o3-mini | 60.4% | 2025-01-31 |
| 13 | Qwen 3 235B | 59.6% | 2025-05-09 |
| 14 | Kimi K2 | 59.1% | 2025-07-17 |
| 15 | Gemini 2.5 Flash | 55.1% | 2025-05-25 |
| 16 | Grok 3 | 53.3% | 2025-04-10 |
| 17 | GPT-4.1 | 52.4% | 2025-04-14 |
| 18 | Claude 3.5 Sonnet | 51.6% | 2025-01-17 |
| 19 | GPT-4o | 45.3% | 2025-03-29 |
| 20 | GPT-4.5 | 44.9% | 2025-02-27 |
| 21 | Gemini 2.0 Pro | 35.6% | 2025-02-25 |
| 22 | o1-mini | 32.9% | 2024-12-22 |
| 23 | Claude 3.5 Haiku | 28.0% | 2024-12-21 |
| 24 | Gemini 2.0 Flash | 22.2% | 2024-12-22 |
| 25 | Llama 4 Maverick | 15.6% | 2025-04-06 |
Upstream leaderboard: https://aider.chat/docs/leaderboards/