Benchmarks
Aider Polyglot
Models scored
72
Best score
88
Metric
Percent correct
Aider Polyglot is an AI evaluation tracked by Epoch AI, with 72 scored model versions on record. The scores here are reported by third parties rather than produced by Epoch AI running the evaluation itself.
Scored models span March 2024 to December 2025. The best score recorded is 88, measured as Percent correct. Models come from OpenAI, Alibaba, DeepSeek, xAI, Moonshot, Google DeepMind, Anthropic, Meta AI and others.
Full record
- Score metric
- Percent correct
- Models scored
- 72
- Best score recorded
- 88
- Earliest model scored
- Mar 26, 2024
- Most recent model scored
- Dec 1, 2025
- Third-party reported
- Yes
Leaderboard
| 01 | gpt-5-2025-08-07_high | OpenAI | 88 |
| 02 | gpt-5-2025-08-07_medium | OpenAI | 86.7 |
| 03 | o3-pro-2025-06-10_high | OpenAI | 84.9 |
| 04 | gemini-2.5-pro-preview-06-05_32K | Google DeepMind | 83.1 |
| 05 | o3-2025-04-16_high | OpenAI | 81.3 |
| 06 | gpt-5-2025-08-07_low | OpenAI | 81.3 |
| 07 | grok-4-0709_high | xAI | 79.6 |
| 08 | grok-4-0709 | xAI | 79.6 |
| 09 | gemini-2.5-pro-preview-06-05 | Google DeepMind | 79.1 |
| 10 | gemini-2.5-pro-preview-05-06 | Google DeepMind | 76.9 |
| 11 | o3-2025-04-16_medium | OpenAI | 76.9 |
| 12 | o3-2025-04-16_unknown | OpenAI | 76.9 |
| 13 | deepseek-reasoner | DeepSeek | 74.2 |
| 14 | DeepSeek-V3.2-Exp_thinking | DeepSeek | 74.2 |
| 15 | gemini-2.5-pro-exp-03-25 | Google DeepMind | 72.9 |
| 16 | gemini-2.5-pro-preview-03-25 | Google DeepMind | 72.9 |
| 17 | claude-opus-4-20250514_32K | Anthropic | 72 |
| 18 | o4-mini-2025-04-16_high | OpenAI | 72 |
| 19 | DeepSeek-R1-0528 | DeepSeek | 71.4 |
| 20 | claude-opus-4-20250514 | Anthropic | 70.7 |
| 21 | DeepSeek-V3.2-Exp | DeepSeek | 70.2 |
| 22 | deepseek-chat | DeepSeek | 70.2 |
| 23 | claude-3-7-sonnet-20250219_32K | Anthropic | 64.9 |
| 24 | o1-2024-12-17_high | OpenAI | 61.7 |
| 25 | claude-sonnet-4-20250514_32K | Anthropic | 61.3 |
| 26 | claude-3-7-sonnet-20250219 | Anthropic | 60.4 |
| 27 | o3-mini-2025-01-31_high | OpenAI | 60.4 |
| 28 | Qwen3-235B-A22B | Alibaba | 59.6 |
| 29 | Qwen3-235B-A22B-Instruct-2507 | Alibaba | 59.6 |
| 30 | moonshotai/kimi-k2-0905 | Moonshot | 59.1 |
| 31 | Kimi-K2-Instruct | Moonshot | 59.1 |
| 32 | DeepSeek-R1 | DeepSeek | 56.9 |
| 33 | claude-sonnet-4-20250514 | Anthropic | 56.4 |
| 34 | gemini-2.5-flash-preview-05-20_23K | Google DeepMind | 55.1 |
| 35 | DeepSeek-V3-0324 | DeepSeek | 55.1 |
| 36 | o3-mini-2025-01-31_medium | OpenAI | 53.8 |
| 37 | grok-3-beta | xAI | 53.3 |
| 38 | gpt-4.1-2025-04-14 | OpenAI | 52.4 |
| 39 | claude-3-5-sonnet-20241022 | Anthropic | 51.6 |
| 40 | grok-3-mini-beta_high | xAI | 49.3 |
| 41 | DeepSeek-V3 | DeepSeek | 48.4 |
| 42 | gemini-2.5-flash-preview-04-17 | Google DeepMind | 47.1 |
| 43 | chatgpt-4o-03-27-2025 | OpenAI | 45.3 |
| 44 | gpt-4.5-preview-2025-02-27 | OpenAI | 44.9 |
| 45 | gemini-2.5-flash-preview-05-20 | Google DeepMind | 44 |
| 46 | openai/gpt-oss-120b_high | OpenAI | 41.8 |
| 47 | gpt-oss-120b_high | OpenAI | 41.8 |
| 48 | Qwen3-32B | Alibaba | 40 |
| 49 | gemini-exp-1206 | Google DeepMind,Google | 38.2 |
| 50 | gemini-2.0-pro-exp-02-05 | Google DeepMind | 35.6 |
| 51 | grok-3-mini-beta_low | xAI | 34.7 |
| 52 | o1-mini-2024-09-12_unknown | OpenAI | 32.9 |
| 53 | gpt-4.1-mini-2025-04-14 | OpenAI | 32.4 |
| 54 | claude-3-5-haiku-20241022 | Anthropic | 28 |
| 55 | chatgpt-4o-01-29-2025 | OpenAI | 27.1 |
| 56 | gpt-4o-2024-08-06 | OpenAI | 23.1 |
| 57 | gemini-2.0-flash-exp | Google DeepMind,Google | 22.2 |
| 58 | qwen-max-2025-01-25 | Alibaba | 21.8 |
| 59 | QwQ-32B | Alibaba | 20.9 |
| 60 | gemini-2.0-flash-thinking-exp-01-21 | Google DeepMind,Google | 18.2 |
More benchmarks
SourceEpoch AI, 'AI Benchmarking Hub'. Published online at epoch.ai. Retrieved 2026-07-29 from https://epoch.ai/benchmarks. Licensed under CC BY 4.0.
← All benchmarks