SimpleBench leaderboard
- Claude Fable 5 leads
- 81.9%
- models measured
- 56
- lead over #2
- +1.3 pts
- median
- 55.9%
Leader on SimpleBench · read October 21
On October 21, 2026, 56 of our catalog's models had a SimpleBench result, and Anthropic Claude Fable 5 led with 81.9%.
Multiple-choice questions on everyday spatial, temporal and social reasoning that people find easy. Share answered correctly.
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 | Claude Fable 5Anthropic | 81.9% | 82.1 | $50.00 | |
| 2 | Claude Opus 5Anthropic | 80.6% | 81.8 | $25.00 | |
| 3 | Gemini 3.1 Pro PreviewGoogle | 79.6% | 35.3 | $12.00 | |
| 4 | GPT-5.5 ProOpenAI | 76.9% | Not benchmarked | $180.00 | |
| 5 | Gemini 3.5 FlashGoogle | 76.7% | 42.8 | $9.00 | |
| 6 | Grok 4.6xAI | 75.9% | 76.4 | $6.00 | |
| 7 | Muse Spark 1.2Meta | 74.5% | 65.1 | $4.25 | |
| 8 | GPT-5.4 ProOpenAI | 74.1% | Not benchmarked | $180.00 | |
| 9 | GPT-5.6 Sol ProOpenAI | 71.7% | Not benchmarked | $20.00 | |
| 10 | Qwen3.7 MaxQwen | 70.4% | 45.1 | $4.425 | |
| 11 | Grok 4.5xAI | 70.0% | 75.5 | $6.00 | |
| 12 | GPT-5.5OpenAI | 69.0% | 67.8 | $30.00 | |
| 13 | Claude Opus 4.6Anthropic | 67.6% | 66.9 | $25.00 | |
| 14 | Claude Opus 4.8Anthropic | 64.8% | 66.0 | $25.00 | |
| 15 | GPT-5.6 SolOpenAI | 64.8% | 76.7 | $20.00 | |
| 16 | Gemini 2.5 Pro Preview 06-05Google | 62.4% | 83.1 | $10.00 | |
| 17 | Claude Opus 4.5Anthropic | 62.0% | 66.8 | $25.00 | |
| 18 | Claude Opus 4.7Anthropic | 61.7% | 51.1 | $25.00 | |
| 19 | GPT-5 ProOpenAI | 61.6% | Not benchmarked | $120.00 | |
| 20 | DeepSeek V4 Flash 0731DeepSeek | 61.1% | Not benchmarked | $1.32 | |
| 21 | Gemini 3 Flash PreviewGoogle | 61.1% | 75.8 | $3.00 | |
| 22 | Kimi K3Moonshot AI | 60.7% | 61.8 | $11.936 | |
| 23 | Claude Sonnet 5Anthropic | 60.6% | 63.5 | $10.00 | |
| 24 | Claude Opus 4.1Anthropic | 60.0% | 27.2 | $75.00 | |
| 25 | GLM 5.2Z.ai | 58.8% | 57.7 | $7.00 | |
| 26 | Kimi K2.7 CodeMoonshot AI | 57.9% | 39.1 | $3.35 | |
| 27 | GPT-5.2 ProOpenAI | 57.4% | Not benchmarked | $168.00 | |
| 28 | GPT-5OpenAI | 56.7% | 72.2 | $10.00 | |
| 29 | GLM 5.1Z.ai | 55.1% | 44.0 | $4.40 | |
| 30 | Claude Sonnet 4.5Anthropic | 54.3% | 59.1 | $15.00 | |
| 31 | GPT-5.1OpenAI | 53.2% | 53.3 | $10.00 | |
| 32 | GLM 5Z.ai | 53.2% | 59.7 | $3.20 | |
| 33 | o3OpenAI | 53.1% | 70.7 | $8.00 | |
| 34 | T InklingThinking Machines | 50.0% | 49.7 | $4.05 | |
| 35 | GPT-5.6 TerraOpenAI | 48.9% | 69.0 | $12.00 | |
| 36 | GLM 4.7Z.ai | 47.7% | 33.6 | $1.98 | |
| 37 | GPT-5.6 LunaOpenAI | 46.8% | 61.5 | $1.20 | |
| 38 | Kimi K2.5Moonshot AI | 46.8% | 58.5 | $2.50 | |
| 39 | DeepSeek V4 Flash 0423DeepSeek | 46.3% | 54.6 | $0.182 | |
| 40 | GPT-5.2OpenAI | 45.8% | 58.9 | $14.00 | |
| 41 | MiniMax M3MiniMax | 45.8% | 57.8 | $1.20 | |
| 42 | Gemini 2.5 FlashGoogle | 41.2% | 28.7 | $2.50 | |
| 43 | R1 0528DeepSeek | 40.8% | 71.4 | $2.15 | |
| 44 | o1OpenAI | 40.1% | 61.7 | $60.00 | |
| 45 | o4 MiniOpenAI | 38.7% | 68.6 | $4.40 | |
| 46 | Qwen3.6 FlashQwen | 35.2% | Not benchmarked | $1.125 | |
| 47 | Llama 4 MaverickMeta Llama | 27.7% | 18.1 | $0.6525 | |
| 48 | DeepSeek V3 0324DeepSeek | 27.2% | 42.0 | $1.14 | |
| 49 | GPT-4.1OpenAI | 27.0% | 46.5 | $8.00 | |
| 50 | Kimi K2 0711Moonshot AI | 26.3% | 64.7 | $2.30 | |
| 51 | o3 MiniOpenAI | 22.8% | 52.1 | $4.40 | |
| 52 | Mistral Large 2407Mistral AI | 22.5% | Not benchmarked | $6.00 | |
| 53 | gpt-oss-120bOpenAI | 22.1% | 34.5 | $0.60 | |
| 54 | Llama 3.3 70B InstructMeta Llama | 19.9% | Not benchmarked | $0.32 | |
| 55 | GPT-4o (2024-08-06)OpenAI | 17.8% | 23.1 | $10.00 | |
| 56 | GPT-4o-mini (2024-07-18)OpenAI | 10.7% | 3.6 | $0.60 |
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.