| GPT-6 Astra | 93.6% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-6 Astra | 93.3% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude Fable 5.1 | 92.9% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-6 Astra | 92.9% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude Fable 5.1 | 92.3% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude Fable 5 | 91.9% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude Opus 5 | 91.8% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude Opus 5 | 91.6% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-5.6 Sol | 89.4% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-5.6 Sol | 88.8% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude Fable 5 | 87.8% | Third party | WeirdML Leaderboard via Epoch AI | — |
| GPT-5.6 Sol | 87.0% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude Opus 5 | 86.3% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-5.5 | 84.9% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-5.5 | 83.9% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude Opus 4.8 | 82.9% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Kimi K3 | 82.6% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-5.3-Codex | 79.3% | Third party | WeirdML Leaderboard via Epoch AI | — |
| GPT-5.6 Terra | 78.3% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude Opus 4.6 | 78.0% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
Claude Opus 4.6Anthropic · Unknown effort | 77.9% | Third party | WeirdML Leaderboard via Epoch AI | — |
| GPT-5.3-Codex | 77.9% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-5.4 | 77.7% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude Opus 4.7 | 76.4% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude Opus 4.7 | 76.4% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
Claude Opus 4.7Anthropic · Unknown effort | 76.4% | Third party | WeirdML Leaderboard via Epoch AI | — |
Claude Opus 4.8Anthropic · Medium effort | 76.0% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude Opus 4.7 | 75.5% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GLM-5.3 | 75.4% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-5.2 | 72.2% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Gemini 3.1 Pro | 72.1% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude Opus 4.8 | 70.5% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GLM-5.2 | 70.1% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Gemini 3 Pro | 69.9% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude Sonnet 5 | 68.8% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GLM-5.2 | 67.3% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
Grok 4.6SpaceXAI (xAI) · High effort | 67.3% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-5.5 | 67.2% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| DeepSeek-V4-Pro | 66.2% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
Claude Sonnet 4.6Anthropic · Medium effort | 66.1% | Third party | WeirdML Leaderboard via Epoch AI | — |
| Claude Opus 4.6 | 65.9% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude Opus 4.5 | 63.7% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-5.2 | 63.4% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| DeepSeek-V4-Flash | 63.0% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Gemini 3.5 Flash | 62.6% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Gemini 3 Flash | 61.6% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-5.6 Luna | 60.9% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-5.1 | 60.8% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-5 | 60.7% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-5 Pro | 60.4% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-5.4 mini | 60.3% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Muse Spark 1.2 | 60.3% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| o3-pro | 58.2% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-5.4 | 57.4% | Third party | WeirdML Leaderboard via Epoch AI | — |
| GPT-5.4 Pro | 57.4% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GLM-5.1 | 57.1% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| DeepSeek-V4-Flash | 57.0% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Gemini 3.6 Flash | 56.1% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Kimi K2.6 | 55.9% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-5-Codex | 54.5% | Third party | WeirdML Leaderboard via Epoch AI | — |
| GPT-5-Codex | 54.5% | Third party | WeirdML Leaderboard via Epoch AI | — |
| Kimi K2.7 Code | 54.1% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Gemini 2.5 Pro | 54.0% | Third party | WeirdML Leaderboard via Epoch AI | — |
| GPT-5 mini | 52.7% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| o4-mini | 52.6% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| o3 | 52.4% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Gemma 4 31B | 52.3% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Grok 4.20 | 52.3% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Gemini 3.1 Flash-Lite | 52.2% | Third party | WeirdML Leaderboard via Epoch AI | — |
Grok 4.3SpaceXAI (xAI) · Unknown effort | 49.9% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-5.2 | 49.6% | Third party | WeirdML Leaderboard via Epoch AI | — |
| GPT-5.2 | 49.6% | Third party | WeirdML Leaderboard via Epoch AI | — |
| GPT-5.4 nano | 49.2% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| DeepSeek-V4-Pro | 48.9% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GLM-5 | 48.2% | Third party | WeirdML Leaderboard via Epoch AI | — |
| gpt-oss-120b | 48.2% | Third party | WeirdML Leaderboard via Epoch AI | — |
| gpt-oss-120b | 48.2% | Third party | WeirdML Leaderboard via Epoch AI | — |
| Claude Sonnet 4.5 | 47.7% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| o1-preview | 47.6% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| DeepSeek-V3.2-Speciale | 46.7% | Third party | WeirdML Leaderboard via Epoch AI | — |
| Claude Sonnet 4.5 | 46.7% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| DeepSeek-V4-Pro | 46.5% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
Grok 4.5SpaceXAI (xAI) · Unknown effort | 46.4% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude Sonnet 4 | 46.1% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| o1 | 46.1% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude Opus 4.1 | 45.9% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Grok 4 | 45.7% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| DeepSeek-V4-Flash | 45.6% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Kimi K2.5 | 45.6% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude Haiku 4.5 | 45.4% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude Haiku 4.5 | 44.1% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude Sonnet 4 | 43.9% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| DeepSeek-V4-Flash | 43.8% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude Opus 4 | 43.7% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Mistral Medium 3.5 | 43.7% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| o3-mini | 43.7% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Mercury 2 | 43.2% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Grok 4 Fast | 42.9% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Kimi K2 Thinking | 42.8% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
Grok 3 MiniSpaceXAI (xAI) · High effort | 42.6% | Third party | WeirdML Leaderboard via Epoch AI | — |
Grok 3 MiniSpaceXAI (xAI) · High effort | 42.6% | Third party | WeirdML Leaderboard via Epoch AI | — |
| Gemini 2.5 Flash | 41.9% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| gpt-oss-120b | 41.9% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| DeepSeek-R1 | 41.6% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Qwen3-Coder-480B-A35B-Instruct | 41.2% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Qwen3-235B-A22B-Thinking-2507 | 41.0% | Third party | WeirdML Leaderboard via Epoch AI | — |
| Gemini 2.5 Flash | 41.0% | Third party | WeirdML Leaderboard via Epoch AI | — |
| gpt-oss-20b | 40.9% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
GLM-4.5Z.ai (Zhipu AI) · Thinking effort | 40.6% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude 3.5 Sonnet | 40.0% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-5 Instant (gpt-5-chat-latest) | 39.8% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Qwen3.5-27B | 39.5% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
DeepSeek-V3.2-ExpDeepSeek · Thinking effort | 39.5% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-4.5 | 39.4% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Kimi K2 | 39.4% | Third party | WeirdML Leaderboard via Epoch AI | — |
| GPT-4.1 | 39.0% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Gemini 3.5 Flash-Lite | 39.0% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Qwen3-235B-A22B-Instruct-2507 | 38.7% | Third party | WeirdML Leaderboard via Epoch AI | — |
| Qwen3-235B-A22B-Thinking-2507 | 38.7% | Third party | WeirdML Leaderboard via Epoch AI | — |
DeepSeek-V3.1DeepSeek · Thinking effort | 38.4% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-5 nano | 38.1% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-5.4 nano | 38.0% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-5.4 mini | 37.9% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-4.1 mini | 37.6% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| DeepSeek-V3.1 | 37.5% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Qwen3-235B-A22B | 37.3% | Third party | WeirdML Leaderboard via Epoch AI | — |
| Qwen3-235B-A22B-Thinking-2507 | 37.3% | Third party | WeirdML Leaderboard via Epoch AI | — |
| Grok 3 | 37.2% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| MiniMax-M2.7 | 37.0% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| gpt-oss-20b | 36.8% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Kimi K2 | 36.7% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| DeepSeek-R1 | 36.5% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| o1-mini | 36.3% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Gemini 2.5 Flash-Lite | 35.2% | Third party | WeirdML Leaderboard via Epoch AI | — |
| Gemma 4 26B A4B | 35.2% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Grok Code Fast 1 | 35.1% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Qwen3.6-35B-A3B | 34.5% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Qwen3-Coder-Next | 34.4% | Third party | WeirdML Leaderboard via Epoch AI | — |
| Mistral Medium 3.1 | 33.1% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Inkling | 32.3% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude 3.5 Sonnet | 31.0% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude 3.5 Haiku | 30.7% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Qwen3-30B-A3B | 29.8% | Third party | WeirdML Leaderboard via Epoch AI | — |
| GPT-5 nano | 25.9% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Gemini 2.0 Flash | 25.8% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-4o | 25.1% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Gemini 1.5 Flash | 24.9% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Llama 4 Maverick | 24.5% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Grok 2 | 22.2% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Gemini 1.5 Pro | 22.2% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Llama 3.1 405B | 21.4% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude 3 Opus | 19.2% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-4.1 nano | 19.0% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-4 Turbo | 18.0% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Qwen2.5-72B-Instruct | 16.0% | Third party | WeirdML Leaderboard via Epoch AI | — |
| Llama 3.3 70B | 14.4% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-4 | 12.4% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-4o mini | 11.8% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Qwen2-72B-Instruct | 11.3% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude 3 Sonnet | 10.2% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude 3 Haiku | 9.8% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Llama 3.1 70B | 9.0% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Claude 2.1 | 7.1% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| GPT-3.5 Turbo | 3.5% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Mixtral 8x22B | 3.2% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |
| Llama 3.1 8B | 1.7% | Third party | https://htihle.github.io/weirdml.html via Epoch AI | — |