SimpleBench
ReportedNo description yet: we only describe an eval set when we can do it correctly.
- Kind
- Published by a third party's leaderboard, collected by Epoch AI. The leaderboard is named and linked.
- Link
- https://simple-bench.com/
- Score shown
- Epoch's column "Score (AVG@5)", as a percentage, higher is better; random guessing scores about 16.7%. The source gives no error bars.
- Results
- 125 in all: 34 for open-weight models, 89 for closed ones, 2 where Epoch does not say. 31 linked to a model page here.
Best open-weight result: deepseek-ai/DeepSeek-V4.1-Flash, 66.7%, number 31 of 125 results.
| # | Model | Weights | Setting | Score | Date | Source |
|---|---|---|---|---|---|---|
| 1 | claude-fable-5-2_unknownAnthropic | closed | effort unknown | 94.9% | not dated | simple-bench.com |
| 2 | gemini-4-carbon_unknownGoogle DeepMind | closed | effort unknown | 92.6% | not dated | simple-bench.com |
| 3 | gemini-4-argon_unknownGoogle DeepMind | closed | effort unknown | 91.7% | not dated | simple-bench.com |
| 4 | claude-opus-5-5_unknownAnthropic | closed | effort unknown | 88.4% | not dated | simple-bench.com |
| 5 | claude-fable-5-1_unknownAnthropic | closed | effort unknown | 86.6% | not dated | simple-bench.com |
| 6 | gpt-6-astra_promaxOpenAI | closed | default | 86.5% | not dated | simple-bench.com |
| 7 | gpt-6-astra_unknownOpenAI | closed | effort unknown | 83.6% | not dated | simple-bench.com |
| 8 | gpt-6.1-sol_unknownOpenAI | closed | effort unknown | 82.9% | not dated | simple-bench.com |
| 9 | gemini-3.8-flash_unknownGoogle DeepMind | closed | effort unknown | 82.4% | not dated | simple-bench.com |
| 10 | claude-fable-5_maxAnthropic | closed | effort max | 81.9% | not dated | simple-bench.com |
| 11 | muse-spark-1.3_unknownMeta AI | closed | effort unknown | 81.8% | not dated | simple-bench.com |
| 12 | claude-opus-5_unknownAnthropic | closed | effort unknown | 80.6% | not dated | lmcouncil.ai |
| 13 | gemini-3.1-pro-previewGoogle DeepMind | closed | default | 79.6% | not dated | lmcouncil.ai |
| 14 | gpt-5.5-pro_highOpenAI | closed | effort high | 76.9% | not dated | SimpleBench Leaderboard, via Epoch AI |
| 15 | gpt-5.5-pro_unknownOpenAI | closed | effort unknown | 76.9% | not dated | lmcouncil.ai |
| 16 | gemini-3.5-flash_unknownGoogle DeepMind | closed | effort unknown | 76.7% | not dated | SimpleBench Leaderboard, via Epoch AI |
| 17 | gemini-3-pro-previewGoogle DeepMind | closed | default | 76.4% | not dated | lmcouncil.ai |
| 18 | claude-sonnet-5-5_unknownAnthropic | closed | effort unknown | 75.9% | not dated | simple-bench.com |
| 19 | grok-4.6_unknownxAI | closed | effort unknown | 75.9% | not dated | lmcouncil.ai |
| 20 | muse-spark-1.2Meta AI | closed | default | 74.5% | not dated | lmcouncil.ai |
| 21 | gpt-5.4-pro-2026-03-05_unknownOpenAI | closed | effort unknown | 74.1% | not dated | SimpleBench Leaderboard |
| 22 | gpt-6-sol_unknownOpenAI | closed | effort unknown | 73.1% | not dated | simple-bench.com |
| 23 | muse-ember-1.0_unknownMeta AI | closed | effort unknown | 72.7% | not dated | simple-bench.com |
| 24 | gpt-6.1-terra_unknownOpenAI | closed | effort unknown | 72.2% | not dated | simple-bench.com |
| 25 | gpt-5.6-sol_prounknownOpenAI | closed | default | 71.7% | not dated | SimpleBench Leaderboard, via Epoch AI |
| 26 | gpt-5.6-sol_proxhighOpenAI | closed | default | 71.7% | not dated | lmcouncil.ai |
| 27 | qwen3.7-maxAlibaba | closed | default | 70.4% | not dated | SimpleBench Leaderboard, via Epoch AI |
| 28 | grok-4.5_unknownxAI | closed | effort unknown | 70.0% | not dated | lmcouncil.ai |
| 29 | gpt-5.5_unknownOpenAI | closed | effort unknown | 69.0% | not dated | lmcouncil.ai |
| 30 | claude-opus-4-6Anthropic | closed | default | 67.6% | not dated | lmcouncil.ai |
| 31 | deepseek-ai/DeepSeek-V4.1-FlashEpoch: deepseek-v4.1-flash_unknown, DeepSeek | open | effort unknown | 66.7% | not dated | simple-bench.com |
| 32 | zai-org/GLM-5.3Epoch: glm-5.3_unknown, Z.ai (Zhipu AI) | open | effort unknown | 66.2% | not dated | simple-bench.com |
| 33 | claude-opus-4-8_unknownAnthropic | closed | effort unknown | 64.8% | not dated | lmcouncil.ai |
| 34 | gpt-5.6-sol_unknownOpenAI | closed | effort unknown | 64.8% | not dated | SimpleBench Leaderboard, via Epoch AI |
| 35 | gpt-5.6-sol_xhighOpenAI | closed | effort xhigh | 64.8% | not dated | lmcouncil.ai |
| 36 | qwen3.6-max-previewAlibaba | not stated | default | 63.0% | not dated | SimpleBench Leaderboard, via Epoch AI |
| 37 | gemini-2.5-pro-preview-06-05Google DeepMind | closed | default | 62.4% | not dated | lmcouncil.ai |
| 38 | claude-opus-4-5-20251101Anthropic | closed | default | 62.0% | not dated | SimpleBench Leaderboard |
| 39 | claude-opus-4-5-20251101_unknownAnthropic | closed | effort unknown | 62.0% | not dated | SimpleBench Leaderboard, via Epoch AI |
| 40 | claude-opus-4-7_unknownAnthropic | closed | effort unknown | 61.7% | not dated | lmcouncil.ai |
| 41 | gpt-5-pro-2025-10-06_highOpenAI | closed | effort high | 61.6% | not dated | SimpleBench Leaderboard |
| 42 | gpt-5-pro-2025-10-06_unknownOpenAI | closed | effort unknown | 61.6% | not dated | SimpleBench Leaderboard, via Epoch AI |
| 43 | deepseek-ai/DeepSeek-V4-Flash-0731Epoch: deepseek-v4-flash-0731_unknown, DeepSeek | open | effort unknown | 61.1% | not dated | lmcouncil.ai |
| 44 | gemini-3-flash-previewGoogle DeepMind | closed | default | 61.1% | not dated | lmcouncil.ai |
| 45 | moonshotai/Kimi-K3Epoch: kimi-k3_max, Moonshot | open | effort max | 60.7% | not dated | lmcouncil.ai |
| 46 | claude-sonnet-5_unknownAnthropic | closed | effort unknown | 60.6% | not dated | SimpleBench Leaderboard, via Epoch AI |
| 47 | grok-4-0709xAI | closed | default | 60.5% | not dated | lmcouncil.ai |
| 48 | Qwen/Qwen3.8-27BEpoch: qwen3.8-27b, Alibaba | open | default | 60.2% | not dated | simple-bench.com |
| 49 | claude-opus-4-1-20250805Anthropic | closed | default | 60.0% | not dated | SimpleBench Leaderboard |
| 50 | claude-opus-4-1-20250805_unknownAnthropic | closed | effort unknown | 60.0% | not dated | SimpleBench Leaderboard, via Epoch AI |
| 51 | claude-haiku-5-5_unknownAnthropic | closed | effort unknown | 59.2% | not dated | simple-bench.com |
| 52 | claude-opus-4-20250514_12KAnthropic | closed | default | 58.8% | not dated | SimpleBench Leaderboard |
| 53 | claude-opus-4-20250514_unknownAnthropic | closed | effort unknown | 58.8% | not dated | SimpleBench Leaderboard, via Epoch AI |
| 54 | zai-org/GLM-5.2Epoch: glm-5.2_unknown, Z.ai (Zhipu AI) | open | effort unknown | 58.8% | not dated | lmcouncil.ai |
| 55 | moonshotai/Kimi-K2.7-CodeEpoch: kimi-k2.7-code, Moonshot | open | default | 57.9% | not dated | lmcouncil.ai |
| 56 | gpt-5.2-pro-2025-12-11OpenAI | closed | default | 57.4% | not dated | lmcouncil.ai |
| 57 | gpt-5-2025-08-07_highOpenAI | closed | effort high | 56.7% | not dated | lmcouncil.ai |
| 58 | grok-4-1-fast-non-reasoningxAI | closed | default | 56.0% | not dated | SimpleBench Leaderboard |
| 59 | grok-4-1-fast-reasoningxAI | closed | default | 56.0% | not dated | SimpleBench Leaderboard, via Epoch AI |
| 60 | zai-org/GLM-5.1Epoch: glm-5.1, Z.ai (Zhipu AI) | open | default | 55.1% | not dated | SimpleBench Leaderboard, via Epoch AI |
| 61 | claude-sonnet-4-5-20250929_12KAnthropic | closed | default | 54.3% | not dated | SimpleBench Leaderboard |
| 62 | claude-sonnet-4-5-20250929_unknownAnthropic | closed | effort unknown | 54.3% | not dated | SimpleBench Leaderboard, via Epoch AI |
| 63 | zai-org/GLM-5Epoch: glm-5, Z.ai (Zhipu AI) | open | default | 53.2% | not dated | SimpleBench Leaderboard, via Epoch AI |
| 64 | gpt-5.1-2025-11-13_highOpenAI | closed | effort high | 53.2% | not dated | lmcouncil.ai |
| 65 | o3-2025-04-16_highOpenAI | closed | effort high | 53.1% | not dated | SimpleBench Leaderboard |
| 66 | deepseek-ai/DeepSeek-V3.2-SpecialeEpoch: DeepSeek-V3.2-Speciale, DeepSeek | open | default | 52.6% | not dated | SimpleBench Leaderboard, via Epoch AI |
| 67 | gemini-2.5-pro-exp-03-25Google DeepMind | closed | default | 51.6% | not dated | SimpleBench Leaderboard |
| 68 | gemini-2.5-pro-preview-03-25Google DeepMind | closed | default | 51.6% | not dated | SimpleBench Leaderboard, via Epoch AI |
| 69 | thinkingmachines/InklingEpoch: Inkling, Thinking Machines | open | default | 50.0% | not dated | lmcouncil.ai |
| 70 | gpt-5.6-terra_unknownOpenAI | closed | effort unknown | 48.9% | not dated | SimpleBench Leaderboard, via Epoch AI |
| 71 | gpt-5.6-terra_xhighOpenAI | closed | effort xhigh | 48.9% | not dated | lmcouncil.ai |
| 72 | zai-org/GLM-4.7Epoch: glm-4.7, Z.ai (Zhipu AI) | open | default | 47.7% | not dated | SimpleBench Leaderboard |
| 73 | zai-org/GLM-4.7Epoch: zai-org/glm-4.7, Z.ai (Zhipu AI) | open | default | 47.7% | not dated | SimpleBench Leaderboard, via Epoch AI |
| 74 | gpt-5.6-luna_unknownOpenAI | closed | effort unknown | 46.8% | not dated | SimpleBench Leaderboard, via Epoch AI |
| 75 | gpt-5.6-luna_xhighOpenAI | closed | effort xhigh | 46.8% | not dated | lmcouncil.ai |
| 76 | moonshotai/Kimi-K2.5Epoch: kimi-k2.5, Moonshot | open | default | 46.8% | not dated | SimpleBench Leaderboard, via Epoch AI |
| 77 | claude-3-7-sonnet-20250219_12KAnthropic | closed | default | 46.4% | not dated | SimpleBench Leaderboard |
| 78 | claude-3-7-sonnet-20250219_unknownAnthropic | closed | effort unknown | 46.4% | not dated | SimpleBench Leaderboard, via Epoch AI |
| 79 | deepseek-ai/DeepSeek-V4-FlashEpoch: deepseek-v4-flash_unknown, DeepSeek | open | effort unknown | 46.3% | not dated | lmcouncil.ai |
| 80 | gpt-5.2-2025-12-11_highOpenAI | closed | effort high | 45.8% | not dated | SimpleBench Leaderboard |
| 81 | gpt-5.2-2025-12-11_unknownOpenAI | closed | effort unknown | 45.8% | not dated | SimpleBench Leaderboard, via Epoch AI |
| 82 | MiniMaxAI/MiniMax-M3Epoch: MiniMax-M3, MiniMax | open | default | 45.8% | not dated | lmcouncil.ai |
| 83 | claude-sonnet-4-20250514_12KAnthropic | closed | default | 45.5% | not dated | SimpleBench Leaderboard |
| 84 | claude-sonnet-4-20250514_unknownAnthropic | closed | effort unknown | 45.5% | not dated | SimpleBench Leaderboard, via Epoch AI |
| 85 | claude-3-7-sonnet-20250219Anthropic | closed | default | 44.9% | not dated | SimpleBench Leaderboard |
| 86 | nemotron-3-ultraNvidiano page match | open | default | 41.7% | not dated | simple-bench.com |
| 87 | o1-preview-2024-09-12OpenAI | closed | default | 41.7% | not dated | SimpleBench Leaderboard |
| 88 | claude-3-5-sonnet-20241022Anthropic | closed | default | 41.4% | not dated | SimpleBench Leaderboard |
| 89 | gemini-2.5-flashGoogle DeepMind | closed | default | 41.2% | not dated | simple-bench.com |
| 90 | deepseek-ai/DeepSeek-R1-0528Epoch: DeepSeek-R1-0528, DeepSeek | open | default | 40.8% | not dated | simple-bench.com |
| 91 | o1-2024-12-17_highOpenAI | closed | effort high | 40.1% | not dated | SimpleBench Leaderboard |
| 92 | deepseek-ai/DeepSeek-V3.1Epoch: DeepSeek-V3.1, DeepSeek | open | default | 40.0% | not dated | simple-bench.com |
| 93 | moonshotai/Kimi-K2-ThinkingEpoch: kimi-k2-thinking, Moonshot | open | default | 39.6% | not dated | simple-bench.com |
| 94 | o4-mini-2025-04-16_highOpenAI | closed | effort high | 38.7% | not dated | SimpleBench Leaderboard |
| 95 | o1-2024-12-17_mediumOpenAI | closed | effort medium | 36.7% | not dated | SimpleBench Leaderboard |
| 96 | grok-3xAI | closed | default | 36.1% | not dated | SimpleBench Leaderboard |
| 97 | qwen3.6-flashAlibaba | not stated | default | 35.2% | not dated | simple-bench.com |
| 98 | MiniMaxAI/MiniMax-M2.1Epoch: MiniMax-M2.1, MiniMax | open | default | 34.7% | not dated | simple-bench.com |
| 99 | gpt-4.5-preview-2025-02-27OpenAI | closed | default | 34.5% | not dated | SimpleBench Leaderboard |
| 100 | gemini-exp-1206Google DeepMind,Google | closed | default | 31.1% | not dated | SimpleBench Leaderboard |
| 101 | Qwen/Qwen3-235B-A22BEpoch: qwen3-235b-a22b, Alibaba | open | default | 31.0% | not dated | SimpleBench Leaderboard |
| 102 | deepseek-ai/DeepSeek-R1Epoch: DeepSeek-R1, DeepSeek | open | default | 30.9% | not dated | SimpleBench Leaderboard |
| 103 | gemini-2.0-flash-thinking-exp-01-21Google DeepMind,Google | closed | default | 30.7% | not dated | SimpleBench Leaderboard |
| 104 | meta-llama/Llama-4-Maverick-17B-128E-InstructEpoch: Llama-4-Maverick-17B-128E-Instruct, Meta AI | open | default | 27.7% | not dated | SimpleBench Leaderboard |
| 105 | claude-3-5-sonnet-20240620Anthropic | closed | default | 27.5% | not dated | SimpleBench Leaderboard |
| 106 | deepseek-ai/DeepSeek-V3-0324Epoch: DeepSeek-V3-0324, DeepSeek | open | default | 27.2% | not dated | SimpleBench Leaderboard |
| 107 | gemini-1.5-pro-002Google DeepMind | closed | default | 27.1% | not dated | SimpleBench Leaderboard |
| 108 | gpt-4.1-2025-04-14OpenAI | closed | default | 27.0% | not dated | SimpleBench Leaderboard |
| 109 | moonshotai/Kimi-K2-InstructEpoch: Kimi-K2-Instruct, Moonshot | open | default | 26.3% | not dated | simple-bench.com |
| 110 | gpt-4-turbo-2024-04-09OpenAI | closed | default | 25.1% | not dated | SimpleBench Leaderboard |
| 111 | MiniMaxAI/MiniMax-M2Epoch: MiniMax-M2, MiniMax | open | default | 25.0% | not dated | simple-bench.com |
| 112 | claude-3-opus-20240229Anthropic | closed | default | 23.5% | not dated | SimpleBench Leaderboard |
| 113 | meta-llama/Llama-3.1-405B-InstructEpoch: Llama-3.1-405B-Instruct, Meta AI | open | default | 23.0% | not dated | SimpleBench Leaderboard |
| 114 | o3-mini-2025-01-31_highOpenAI | closed | effort high | 22.8% | not dated | SimpleBench Leaderboard |
| 115 | grok-2-1212xAI | closed | default | 22.7% | not dated | SimpleBench Leaderboard |
| 116 | mistral-large-2407Mistral AIno page match | open | default | 22.5% | not dated | SimpleBench Leaderboard |
| 117 | openai/gpt-oss-120bEpoch: gpt-oss-120b, OpenAI | open | default | 22.1% | not dated | simple-bench.com |
| 118 | mistral-large-2512Mistral AIno page match | open | default | 20.4% | not dated | simple-bench.com |
| 119 | meta-llama/Llama-3.3-70B-InstructEpoch: Llama-3.3-70B-Instruct, Meta AI | open | default | 19.9% | not dated | SimpleBench Leaderboard |
| 120 | deepseek-ai/DeepSeek-V3Epoch: DeepSeek-V3, DeepSeek | open | default | 18.9% | not dated | SimpleBench Leaderboard |
| 121 | gemini-2.0-flash-expGoogle DeepMind,Google | closed | default | 18.9% | not dated | SimpleBench Leaderboard |
| 122 | o1-mini-2024-09-12_mediumOpenAI | closed | effort medium | 18.1% | not dated | SimpleBench Leaderboard |
| 123 | gpt-4o-2024-08-06OpenAI | closed | default | 17.8% | not dated | SimpleBench Leaderboard |
| 124 | CohereLabs/c4ai-command-r-plus-08-2024Epoch: c4ai-command-r-plus-08-2024, Cohere,Cohere Labs (formerly Cohere for AI) | open | default | 17.4% | not dated | SimpleBench Leaderboard |
| 125 | gpt-4o-mini-2024-07-18OpenAI | closed | default | 10.7% | not dated | SimpleBench Leaderboard |
Open-weight rows are shaded. Dates are when the result was run or evaluated, where the data says. Data from Epoch AI (CC-BY 4.0), fetched 19 Nov 2026: Epoch AI, ‘Capabilities & benchmarking’. Published online at epoch.ai. Retrieved from ‘https://epoch.ai/benchmarks’ [online resource]. Epoch AI. Method