Calibration
AA-Omniscience Accuracy → BioMysteryBench (human-difficult)
BioMysteryBench (human-difficult) is estimated from AA-Omniscience Accuracy with a Michaelis–Menten curve fitted on 6 models measured on both: y = 0.8393·x / (0.42005 + x), R² = 0.58, cross-validated error 5.9 pp. It is used for 177 estimates.
| Estimated model | AA-Omniscience Accuracy | BioMysteryBench (human-difficult) | Source |
|---|---|---|---|
| A.X K2 | 18.6% | 25.8% | estimated ± 5.9 pp, low confidence |
| Apodex 1.1 Mini | 31.7% | 36.1% | estimated ± 5.9 pp, low confidence |
| Celeris-1 | 11.0% | 17.4% | estimated ± 5.9 pp, low confidence |
| Claude 3 Haiku | 17.6% | 24.8% | estimated ± 5.9 pp, low confidence |
| Claude 4 Sonnet | 22.7% | 29.4% | estimated ± 5.9 pp, low confidence |
| Claude Fable 5 | 65.4% | 51.1% | estimated ± 5.9 pp, low confidence |
| Claude Fable 5.1 | 67.2% | 51.6% | estimated ± 5.9 pp, low confidence |
| Claude Opus 4.5 | 40.9% | 41.4% | estimated ± 5.9 pp, low confidence |
| Claude Opus 4.5 Thinking | 46.6% | 44.1% | estimated ± 5.9 pp, low confidence |
| Claude Opus 4.6 | 45.8% | 43.8% | estimated ± 5.9 pp, low confidence |
| Claude Opus 4.6 (Adaptive) | 47.0% | 44.3% | estimated ± 5.9 pp, low confidence |
| Claude Opus 4.7 | 44.7% | 43.3% | estimated ± 5.9 pp, low confidence |
| Claude Opus 4.7 (Adaptive) | 48.9% | 45.1% | estimated ± 5.9 pp, low confidence |
| Claude Opus 4.8 | 48.8% | 45.1% | estimated ± 5.9 pp, low confidence |
| Claude Sonnet 4.6 | 38.6% | 40.2% | estimated ± 5.9 pp, low confidence |
| Claude Sonnet 5 | 40.1% | 41.0% | estimated ± 5.9 pp, low confidence |
| Command A+ | 8.9% | 14.7% | estimated ± 5.9 pp, low confidence |
| DeepSeek-R1 | 30.5% | 35.3% | estimated ± 5.9 pp, low confidence |
| DeepSeek V3 | 25.5% | 31.7% | estimated ± 5.9 pp, low confidence |
| DeepSeek V3 0324 | 24.3% | 30.8% | estimated ± 5.9 pp, low confidence |
| DeepSeek V3.1 | 23.1% | 29.8% | estimated ± 5.9 pp, low confidence |
| DeepSeek V3.1 (Reasoning) | 29.0% | 34.3% | estimated ± 5.9 pp, low confidence |
| DeepSeek V3.2 | 24.0% | 30.5% | estimated ± 5.9 pp, low confidence |
| DeepSeek V4.1 Flash | 46.4% | 44.1% | estimated ± 5.9 pp, low confidence |
| DeepSeek V4 Flash 0731 | 40.4% | 41.1% | estimated ± 5.9 pp, low confidence |
| DeepSeek V4 Pro 0813 | 49.1% | 45.2% | estimated ± 5.9 pp, low confidence |
| Exaone 4.0 1.2B | 5.0% | 8.9% | estimated ± 5.9 pp, low confidence |
| Exaone 4.0 32B | 10.6% | 16.9% | estimated ± 5.9 pp, low confidence |
| Gemini 2.5 Flash | 26.1% | 32.2% | estimated ± 5.9 pp, low confidence |
| Gemini 2.5 Pro | 39.1% | 40.5% | estimated ± 5.9 pp, low confidence |
| Gemini 3.1 Pro | 54.9% | 47.5% | estimated ± 5.9 pp, low confidence |
| Gemini 3.5 Flash | 51.9% | 46.4% | estimated ± 5.9 pp, low confidence |
| Gemini 3.5 Flash-Lite | 29.5% | 34.6% | estimated ± 5.9 pp, low confidence |
| Gemini 3.6 Flash | 50.0% | 45.6% | estimated ± 5.9 pp, low confidence |
| Gemini 3 Flash | 45.8% | 43.8% | estimated ± 5.9 pp, low confidence |
| Gemini 3 Pro | 55.8% | 47.9% | estimated ± 5.9 pp, low confidence |
| Gemini 4 Argon | 49.9% | 45.6% | estimated ± 5.9 pp, low confidence |
| Gemma 3 27B | 13.0% | 19.8% | estimated ± 5.9 pp, low confidence |
| Gemma 4 12B | 15.6% | 22.7% | estimated ± 5.9 pp, low confidence |
| Gemma 4 26B A4B | 19.1% | 26.2% | estimated ± 5.9 pp, low confidence |
| Gemma 4 31B | 20.0% | 27.1% | estimated ± 5.9 pp, low confidence |
| Gemma 4 E2B | 6.6% | 11.4% | estimated ± 5.9 pp, low confidence |
| Gemma 4 E4B | 8.6% | 14.3% | estimated ± 5.9 pp, low confidence |
| GLM-4.5-Air | 16.3% | 23.5% | estimated ± 5.9 pp, low confidence |
| GLM-4.6 | 21.4% | 28.3% | estimated ± 5.9 pp, low confidence |
| GLM-4.7 | 29.3% | 34.5% | estimated ± 5.9 pp, low confidence |
| GLM-5 | 26.3% | 32.3% | estimated ± 5.9 pp, low confidence |
| GLM-5.1 | 23.7% | 30.3% | estimated ± 5.9 pp, low confidence |
| GLM-5.2 | 24.3% | 30.8% | estimated ± 5.9 pp, low confidence |
| GLM-5.3 | 33.9% | 37.5% | estimated ± 5.9 pp, low confidence |
| GLM-5-Turbo | 28.4% | 33.9% | estimated ± 5.9 pp, low confidence |
| GLM-5V-Turbo | 29.3% | 34.5% | estimated ± 5.9 pp, low confidence |
| GPT-4.1 | 27.8% | 33.4% | estimated ± 5.9 pp, low confidence |
| GPT-4.1 mini | 20.3% | 27.3% | estimated ± 5.9 pp, low confidence |
| GPT-4.1 nano | 13.7% | 20.6% | estimated ± 5.9 pp, low confidence |
| GPT-4o | 19.9% | 27.0% | estimated ± 5.9 pp, low confidence |
| GPT-5.1 | 37.7% | 39.7% | estimated ± 5.9 pp, low confidence |
| GPT-5.1-Codex | 39.9% | 40.9% | estimated ± 5.9 pp, low confidence |
| GPT-5.1-Codex-Max | 39.9% | 40.9% | estimated ± 5.9 pp, low confidence |
| GPT-5.2 | 44.3% | 43.1% | estimated ± 5.9 pp, low confidence |
| GPT-5.2-Codex | 41.1% | 41.5% | estimated ± 5.9 pp, low confidence |
| GPT-5.3 Codex | 52.9% | 46.8% | estimated ± 5.9 pp, low confidence |
| GPT-5.4 | 50.8% | 45.9% | estimated ± 5.9 pp, low confidence |
| GPT-5.4 mini | 37.5% | 39.6% | estimated ± 5.9 pp, low confidence |
| GPT-5.4 nano | 25.7% | 31.9% | estimated ± 5.9 pp, low confidence |
| GPT-5.5 | 58.0% | 48.7% | estimated ± 5.9 pp, low confidence |
| GPT-5.6 Luna | 42.7% | 42.3% | estimated ± 5.9 pp, low confidence |
| GPT-5.6 Sol | 59.4% | 49.2% | estimated ± 5.9 pp, low confidence |
| GPT-5.6 Terra | 46.8% | 44.2% | estimated ± 5.9 pp, low confidence |
| GPT-5 (high) | 40.3% | 41.1% | estimated ± 5.9 pp, low confidence |
| GPT-5 (medium) | 39.5% | 40.7% | estimated ± 5.9 pp, low confidence |
| GPT-6.1 Sol | 62.1% | 50.1% | estimated ± 5.9 pp, low confidence |
| GPT-6 Astra | 62.6% | 50.2% | estimated ± 5.9 pp, low confidence |
| GPT-6 Luna | 43.8% | 42.8% | estimated ± 5.9 pp, low confidence |
| GPT-6 Sol | 54.5% | 47.4% | estimated ± 5.9 pp, low confidence |
| GPT-OSS 120B | 21.8% | 28.7% | estimated ± 5.9 pp, low confidence |
| GPT-OSS 20B | 16.0% | 23.2% | estimated ± 5.9 pp, low confidence |
| Granite-4.0-350M | 3.9% | 7.1% | estimated ± 5.9 pp, low confidence |
| Granite-4.0-H-1B | 5.2% | 9.2% | estimated ± 5.9 pp, low confidence |
| Granite-4.0-H-350M | 3.8% | 7.0% | estimated ± 5.9 pp, low confidence |
| Granite 4.2 30B | 10.1% | 16.3% | estimated ± 5.9 pp, low confidence |
| Granite 4.2 3B | 9.2% | 15.1% | estimated ± 5.9 pp, low confidence |
| Granite 4.2 8B | 11.2% | 17.7% | estimated ± 5.9 pp, low confidence |
| Grok 4 | 40.5% | 41.2% | estimated ± 5.9 pp, low confidence |
| Grok 4.1 Fast | 17.2% | 24.4% | estimated ± 5.9 pp, low confidence |
| Grok 4.1 Fast (Reasoning) | 25.1% | 31.4% | estimated ± 5.9 pp, low confidence |
| Grok 4.3 | 34.6% | 37.9% | estimated ± 5.9 pp, low confidence |
| Grok 4.5 | 51.6% | 46.3% | estimated ± 5.9 pp, low confidence |
| Grok 4.6 | 48.2% | 44.8% | estimated ± 5.9 pp, low confidence |
| Grok 4.7 | 47.4% | 44.5% | estimated ± 5.9 pp, low confidence |
| Grok 4 Fast (Reasoning) | 22.8% | 29.5% | estimated ± 5.9 pp, low confidence |
| Grok Code Fast 1 | 23.5% | 30.1% | estimated ± 5.9 pp, low confidence |
| Hy3 | 32.0% | 36.3% | estimated ± 5.9 pp, low confidence |
| Hy3 Preview | 31.5% | 36.0% | estimated ± 5.9 pp, low confidence |
| Inkling | 41.6% | 41.8% | estimated ± 5.9 pp, low confidence |
| Inkling-Small | 33.2% | 37.1% | estimated ± 5.9 pp, low confidence |
| K-Exaone | 16.4% | 23.6% | estimated ± 5.9 pp, low confidence |
| K-EXAONE 2.0 | 13.1% | 20.0% | estimated ± 5.9 pp, low confidence |
| Kimi K2.6 | 32.6% | 36.7% | estimated ± 5.9 pp, low confidence |
| Kimi K2 | 27.4% | 33.1% | estimated ± 5.9 pp, low confidence |
| Kimi K2.5 | 35.2% | 38.3% | estimated ± 5.9 pp, low confidence |
| Kimi K2.5 (Reasoning) | 35.2% | 38.3% | estimated ± 5.9 pp, low confidence |
| Kimi K2.7 Code | 39.6% | 40.7% | estimated ± 5.9 pp, low confidence |
| Kimi K3 | 47.6% | 44.6% | estimated ± 5.9 pp, low confidence |
| LFM2.5-2.6B | 4.4% | 8.0% | estimated ± 5.9 pp, low confidence |
| LFM2.5-8B-A1B | 9.4% | 15.3% | estimated ± 5.9 pp, low confidence |
| LFM2.5-VL-1.6B-Extract | 5.8% | 10.2% | estimated ± 5.9 pp, low confidence |
| Ling 2.6 Flash | 15.6% | 22.7% | estimated ± 5.9 pp, low confidence |
| Ling 3.0 Flash | 18.2% | 25.4% | estimated ± 5.9 pp, low confidence |
| Ling 3.0 Flash FP8 | 18.2% | 25.4% | estimated ± 5.9 pp, low confidence |
| Ling 3.0 Flash VL | 14.4% | 21.4% | estimated ± 5.9 pp, low confidence |
| Ling 3.0 Tiny | 8.5% | 14.1% | estimated ± 5.9 pp, low confidence |
| Ling 3.1 Flash | 29.1% | 34.3% | estimated ± 5.9 pp, low confidence |
| Llama 3.1 405B | 23.2% | 29.9% | estimated ± 5.9 pp, low confidence |
| Llama 4 Maverick | 24.9% | 31.2% | estimated ± 5.9 pp, low confidence |
| Llama 4 Scout | 15.2% | 22.3% | estimated ± 5.9 pp, low confidence |
| Mercury 2.5 | 22.0% | 28.8% | estimated ± 5.9 pp, low confidence |
| MiMo-V2.5-Pro | 22.4% | 29.2% | estimated ± 5.9 pp, low confidence |
| MiMo-V2.6-Flash | 27.0% | 32.8% | estimated ± 5.9 pp, low confidence |
| MiMo-V2.6-Pro | 34.8% | 38.0% | estimated ± 5.9 pp, low confidence |
| MiMo-V2-Flash | 15.6% | 22.7% | estimated ± 5.9 pp, low confidence |
| MiMo-V2-Omni | 19.3% | 26.4% | estimated ± 5.9 pp, low confidence |
| MiMo-V2-Pro | 26.6% | 32.5% | estimated ± 5.9 pp, low confidence |
| MiniCPM5-2B | 8.4% | 14.0% | estimated ± 5.9 pp, low confidence |
| MiniMax M2.7 | 26.8% | 32.7% | estimated ± 5.9 pp, low confidence |
| MiniMax M3 | 16.7% | 23.9% | estimated ± 5.9 pp, low confidence |
| Mistral Large 2 | 19.9% | 27.0% | estimated ± 5.9 pp, low confidence |
| Mistral Large 3 | 25.0% | 31.3% | estimated ± 5.9 pp, low confidence |
| Mistral Large 4 | 25.8% | 31.9% | estimated ± 5.9 pp, low confidence |
| Mistral Medium 3 | 18.3% | 25.5% | estimated ± 5.9 pp, low confidence |
| Mistral Medium 3.5 128B | 24.7% | 31.1% | estimated ± 5.9 pp, low confidence |
| Mistral Small 4 | 21.7% | 28.6% | estimated ± 5.9 pp, low confidence |
| Mistral Small 4 (Reasoning) | 21.7% | 28.6% | estimated ± 5.9 pp, low confidence |
| Muse Glimmer 30B | 27.0% | 32.8% | estimated ± 5.9 pp, low confidence |
| Muse Spark | 49.6% | 45.4% | estimated ± 5.9 pp, low confidence |
| Muse Spark 1.1 | 52.1% | 46.5% | estimated ± 5.9 pp, low confidence |
| Muse Spark 1.2 | 45.4% | 43.6% | estimated ± 5.9 pp, low confidence |
| Muse Spark 1.3 | 43.6% | 42.7% | estimated ± 5.9 pp, low confidence |
| Nemotron 3.5 Lightning 30B A3B NVFP4 | 14.4% | 21.4% | estimated ± 5.9 pp, low confidence |
| Nemotron 3 Nano 30B | 17.3% | 24.5% | estimated ± 5.9 pp, low confidence |
| Nemotron 3 Nano Omni 30B A3B | 15.2% | 22.3% | estimated ± 5.9 pp, low confidence |
| Nemotron 3 Super 100B | 24.3% | 30.8% | estimated ± 5.9 pp, low confidence |
| Nemotron 3 Ultra | 21.6% | 28.5% | estimated ± 5.9 pp, low confidence |
| Nemotron Ultra 253B | 20.1% | 27.2% | estimated ± 5.9 pp, low confidence |
| North Mini Code | 18.9% | 26.0% | estimated ± 5.9 pp, low confidence |
| Nova Pro | 16.9% | 24.1% | estimated ± 5.9 pp, low confidence |
| o1 | 34.5% | 37.8% | estimated ± 5.9 pp, low confidence |
| o3 | 38.6% | 40.2% | estimated ± 5.9 pp, low confidence |
| Phi-4 | 14.1% | 21.1% | estimated ± 5.9 pp, low confidence |
| Quasar 438B | 15.5% | 22.6% | estimated ± 5.9 pp, low confidence |
| Qwen3.5-122B-A10B | 24.4% | 30.8% | estimated ± 5.9 pp, low confidence |
| Qwen3.5-27B | 20.7% | 27.7% | estimated ± 5.9 pp, low confidence |
| Qwen3.5-35B-A3B | 20.1% | 27.2% | estimated ± 5.9 pp, low confidence |
| Qwen3.5 397B | 24.5% | 30.9% | estimated ± 5.9 pp, low confidence |
| Qwen3.5 397B (Reasoning) | 24.5% | 30.9% | estimated ± 5.9 pp, low confidence |
| Qwen3.6-27B | 19.6% | 26.7% | estimated ± 5.9 pp, low confidence |
| Qwen3.6-35B-A3B | 18.8% | 25.9% | estimated ± 5.9 pp, low confidence |
| Qwen 3.6 Max (preview) | 37.9% | 39.8% | estimated ± 5.9 pp, low confidence |
| Qwen3.6 Plus | 26.4% | 32.4% | estimated ± 5.9 pp, low confidence |
| Qwen3.7 Max | 31.1% | 35.7% | estimated ± 5.9 pp, low confidence |
| Qwen3.7 Plus | 22.5% | 29.3% | estimated ± 5.9 pp, low confidence |
| Qwen3.8-27B | 15.6% | 22.7% | estimated ± 5.9 pp, low confidence |
| Qwen3.8-Flash-Next | 24.5% | 30.9% | estimated ± 5.9 pp, low confidence |
| Qwen3.8 Max Preview | 31.7% | 36.1% | estimated ± 5.9 pp, low confidence |
| Qwen3 Max | 24.4% | 30.8% | estimated ± 5.9 pp, low confidence |
| Qwen3-Omni-30B-A3B-Instruct | 14.3% | 21.3% | estimated ± 5.9 pp, low confidence |
| Qwen3-Omni-30B-A3B-Thinking | 14.6% | 21.6% | estimated ± 5.9 pp, low confidence |
| Sarvam 105B | 17.6% | 24.8% | estimated ± 5.9 pp, low confidence |
| Sarvam 30B | 12.6% | 19.4% | estimated ± 5.9 pp, low confidence |
| Solar Pro 2 | 16.1% | 23.3% | estimated ± 5.9 pp, low confidence |
| Solar Pro 3 | 18.5% | 25.7% | estimated ± 5.9 pp, low confidence |
| Solar Pro 4 | 18.9% | 26.0% | estimated ± 5.9 pp, low confidence |
| Step 3.7 Flash | 25.8% | 31.9% | estimated ± 5.9 pp, low confidence |
| Step 5 Preview | 41.5% | 41.7% | estimated ± 5.9 pp, low confidence |
| Trinity-Large-Preview | 22.5% | 29.3% | estimated ± 5.9 pp, low confidence |
| Trinity-Large-Thinking | 22.5% | 29.3% | estimated ± 5.9 pp, low confidence |
| Ultravox v0.6 Llama 3.3 70B | 19.0% | 26.1% | estimated ± 5.9 pp, low confidence |