Chess puzzles leaderboard
- GPT-6 Astra Pro leads
- 76.0%
- models measured
- 81
- lead over #2
- +2.0 pts
- median
- 37.0%
Leader on Chess puzzles · read October 21
On October 21, 2026, 81 of our catalog's models had a Chess puzzles result, and OpenAI GPT-6 Astra Pro led with 76.0%.
Epoch AI collects its results; its benchmarking hub describes the tasks and links the authors.
Role in the index
We collect this benchmark and show it as supporting evidence. It is not part of the Vibe Coding Index, because it measures something other than coding and the index is about coding. Read the methodology
| # | Model | Result | Evidence | Coding index | Output $/Mtok |
|---|---|---|---|---|---|
| 1 | GPT-6 Astra ProOpenAI | 76.0% | Not benchmarked | $50.00 | |
| 2 | GPT-6 AstraOpenAI | 74.0% | 88.7 | $50.00 | |
| 3 | GPT-5.6 Sol ProOpenAI | 68.0% | Not benchmarked | $20.00 | |
| 4 | GPT-6.1 SolOpenAI | 68.0% | 88.7 | $10.00 | |
| 5 | Claude Opus 5.5Anthropic | 67.0% | 92.2 | $20.00 | |
| 6 | Gemini 3.7 FlashGoogle | 66.0% | 69.9 | $3.75 | |
| 7 | GPT-5.5OpenAI | 63.0% | 67.8 | $30.00 | |
| 8 | GPT-5.5 ProOpenAI | 62.0% | Not benchmarked | $180.00 | |
| 9 | GPT-5.2 ProOpenAI | 60.0% | Not benchmarked | $168.00 | |
| 10 | Claude Sonnet 5.5Anthropic | 59.0% | 88.9 | $10.00 | |
| 11 | Gemini 4 ArgonGoogle | 59.0% | 80.5 | $10.00 | |
| 12 | GPT-5.4 ProOpenAI | 59.0% | Not benchmarked | $180.00 | |
| 13 | GPT-5.6 SolOpenAI | 59.0% | 76.7 | $20.00 | |
| 14 | Gemini 3.8 FlashGoogle | 58.0% | 70.5 | $4.125 | |
| 15 | GPT-6 SolOpenAI | 56.0% | 82.8 | $10.00 | |
| 16 | Claude Haiku 5.5Anthropic | 55.0% | 75.9 | $0.50 | |
| 17 | Claude Opus 4.8Anthropic | 52.0% | 66.0 | $25.00 | |
| 18 | Gemini 3.1 Pro PreviewGoogle | 51.0% | 35.3 | $12.00 | |
| 19 | Claude Fable 5.1Anthropic | 48.0% | 89.1 | $50.00 | |
| 20 | DeepSeek V4.1 FlashDeepSeek | 48.0% | 77.2 | $0.525 | |
| 21 | GPT-5 ProOpenAI | 48.0% | Not benchmarked | $120.00 | |
| 22 | GPT-5.2OpenAI | 48.0% | 58.9 | $14.00 | |
| 23 | GPT-5.4OpenAI | 48.0% | 37.7 | $15.00 | |
| 24 | Claude Fable 5Anthropic | 47.0% | 82.1 | $50.00 | |
| 25 | GPT-5.6 TerraOpenAI | 47.0% | 69.0 | $12.00 | |
| 26 | DeepSeek V4 Pro 0813DeepSeek | 46.0% | 54.6 | $1.98 | |
| 27 | Qwen3.8 Max (0902)Qwen | 46.0% | 63.5 | $6.00 | |
| 28 | Claude Sonnet 5Anthropic | 45.0% | 63.5 | $10.00 | |
| 29 | Gemini 3.5 FlashGoogle | 45.0% | 42.8 | $9.00 | |
| 30 | GPT-5.4 NanoOpenAI | 45.0% | Not benchmarked | $1.25 | |
| 31 | Claude Opus 5Anthropic | 43.0% | 81.8 | $25.00 | |
| 32 | GPT-5.6 LunaOpenAI | 43.0% | 61.5 | $1.20 | |
| 33 | Gemini 3.6 FlashGoogle | 42.0% | 48.3 | $3.75 | |
| 34 | GPT-5.1OpenAI | 42.0% | 53.3 | $10.00 | |
| 35 | Claude Opus 4.7Anthropic | 41.0% | 51.1 | $25.00 | |
| 36 | Grok 4.7xAI | 41.0% | 75.0 | $6.00 | |
| 37 | Gemini 3 Flash PreviewGoogle | 39.0% | 75.8 | $3.00 | |
| 38 | Muse Spark 1.3Meta | 37.0% | 79.7 | $4.25 | |
| 39 | Kimi K3Moonshot AI | 37.0% | 61.8 | $11.936 | |
| 40 | o3OpenAI | 37.0% | 70.7 | $8.00 | |
| 41 | GLM 5.3Z.ai | 37.0% | 73.1 | $1.40 | |
| 42 | GPT-5OpenAI | 36.0% | 72.2 | $10.00 | |
| 43 | GPT-6 LunaOpenAI | 36.0% | 71.4 | $0.50 | |
| 44 | Grok 4.5xAI | 35.0% | 75.5 | $6.00 | |
| 45 | Muse Spark 1.1Meta | 34.0% | 48.9 | $4.25 | |
| 46 | Mistral Large 4Mistral AI | 33.0% | 55.3 | $4.18 | |
| 47 | Grok 4.6xAI | 33.0% | 76.4 | $6.00 | |
| 48 | DeepSeek V4 Flash 0731DeepSeek | 32.0% | Not benchmarked | $1.32 | |
| 49 | Muse Spark 1.2Meta | 27.0% | 65.1 | $4.25 | |
| 50 | Qwen3.7 MaxQwen | 26.0% | 45.1 | $4.425 | |
| 51 | GPT-5 MiniOpenAI | 24.0% | 59.8 | $2.00 | |
| 52 | Claude Sonnet 4.6Anthropic | 23.0% | 57.8 | $15.00 | |
| 53 | GPT-5.4 MiniOpenAI | 23.0% | 28.2 | $4.50 | |
| 54 | GPT-5 NanoOpenAI | 22.0% | 34.8 | $0.40 | |
| 55 | o4 MiniOpenAI | 22.0% | 68.6 | $4.40 | |
| 56 | GLM 5.2Z.ai | 22.0% | 57.7 | $7.00 | |
| 57 | Claude Opus 4.6Anthropic | 21.0% | 66.9 | $25.00 | |
| 58 | Gemini 3.5 Flash LiteGoogle | 21.0% | 33.6 | $2.50 | |
| 59 | T InklingThinking Machines | 21.0% | 49.7 | $4.05 | |
| 60 | GLM 4.7Z.ai | 21.0% | 33.6 | $1.98 | |
| 61 | GLM 5.3 FlashZ.ai | 21.0% | 71.0 | $0.50 | |
| 62 | Kimi K2.6Moonshot AI | 20.0% | 44.2 | $3.41 | |
| 63 | GLM 5.1Z.ai | 20.0% | 44.0 | $4.40 | |
| 64 | DeepSeek V4 Pro 0423DeepSeek | 19.0% | 35.3 | $1.9105 | |
| 65 | Grok 4.3xAI | 18.0% | 22.5 | $2.50 | |
| 66 | o3 MiniOpenAI | 15.0% | 52.1 | $4.40 | |
| 67 | Claude Opus 4.5Anthropic | 14.0% | 66.8 | $25.00 | |
| 68 | Kimi K2.7 CodeMoonshot AI | 14.0% | 39.1 | $3.35 | |
| 69 | Qwen3.5-FlashQwen | 14.0% | 6.2 | $0.26 | |
| 70 | Qwen3.6 PlusQwen | 14.0% | 37.4 | $1.95 | |
| 71 | MiniMax M3MiniMax | 13.0% | 57.8 | $1.20 | |
| 72 | Qwen3.6 FlashQwen | 12.0% | Not benchmarked | $1.125 | |
| 73 | Claude Sonnet 4.5Anthropic | 10.0% | 59.1 | $15.00 | |
| 74 | o1OpenAI | 10.0% | 61.7 | $60.00 | |
| 75 | Claude Haiku 4.5Anthropic | 8.0% | 49.8 | $5.00 | |
| 76 | Qwen3.5 Plus 2026-04-20Qwen | 8.0% | Not benchmarked | $1.80 | |
| 77 | GPT-4o (2024-08-06)OpenAI | 6.0% | 23.1 | $10.00 | |
| 78 | GPT-4.1OpenAI | 3.0% | 46.5 | $8.00 | |
| 79 | GPT-4o-mini (2024-07-18)OpenAI | 3.0% | 3.6 | $0.60 | |
| 80 | GPT-4.1 MiniOpenAI | 1.0% | 28.5 | $1.60 | |
| 81 | GPT-4.1 NanoOpenAI | 0.0% | 8.9 | $0.40 |
Only models in our catalog are listed, and each appears once with its best published run. 197 rows were read from the source on October 21, 2026, and rows that match no catalog model are left out.