Health Evals

Coverage

148 of 483 model and board pairs have a published result.

Model by board grid. Cells give the rank within that board.
ModelHealthBench ProfessionalHealthBenchHealth Optimization BenchHealthBench HardWHBenchArtificial Analysis Healthcare & Medical IndexHealth Optimization Bench: subject suites
GPT-5.6 Sol10 / 3115 / 286 / 186 / 199 / 294 / 16
Claude Fable 53 / 315 / 282 / 184 / 291 / 16
Claude Fable 5.17 / 318 / 281 / 182 / 298 / 16
Claude Opus 511 / 311 / 285 / 183 / 292 / 16
GPT-6 Astra1 / 3111 / 283 / 183 / 194 / 29
Claude Sonnet 513 / 3110 / 2813 / 1811 / 16
GPT-5.424 / 3123 / 2814 / 193 / 22
GPT-5.6 Luna18 / 3119 / 288 / 1921 / 29
GPT-5.6 Terra14 / 3115 / 287 / 1919 / 29
GPT-6 Luna8 / 3121 / 2810 / 1920 / 29
GPT-6 Sol8 / 3125 / 2812 / 1913 / 29
Grok 4.623 / 314 / 1812 / 293 / 16
Muse Spark19 / 319 / 181 / 197 / 16
Claude Opus 5.54 / 316 / 281 / 29
GPT OSS 120B14 / 2813 / 1928 / 29
GPT-525 / 3113 / 284 / 19
GPT-5.129 / 3127 / 2817 / 19
GPT-5.226 / 314 / 282 / 19
GPT-5.522 / 3118 / 289 / 19
GPT-5.5 Instant30 / 3126 / 2818 / 19
Inkling12 / 1824 / 2910 / 16
Kimi K37 / 189 / 296 / 16
MAI-Thinking-131 / 3115 / 1814 / 16
MiniMax M314 / 1823 / 2913 / 16
Mistral Medium 3.517 / 1828 / 2915 / 16
Claude Opus 4.815 / 319 / 28
Claude Sonnet 4.627 / 312 / 22
Claude Sonnet 5.52 / 312 / 28
Gemini 3.611 / 189 / 16
Gemini 3.8 Flash10 / 1814 / 29
GLM 5.216 / 1812 / 16
GPT OSS 20B28 / 2819 / 19
GPT-5.3 Chat22 / 2816 / 19
Grok 4.716 / 317 / 29
Nemotron 3.5 Lightning18 / 1816 / 16
Baichuan-M33 / 28
Claude Opus 414 / 22
Claude Opus 4.61 / 22
Claude Opus 4.721 / 31
Claude Sonnet 415 / 22
DeepSeek R110 / 22
DeepSeek V4.1 Flash16 / 29
DeepSeek-V3.26 / 22
Gemini 2.5 Flash13 / 22
Gemini 2.5 Pro21 / 22
Gemini 3 Flash4 / 22
Gemini 3.5 Flash Lite25 / 29
GLM 5.37 / 29
GLM 5.3 Flash9 / 29
GPT-4.111 / 22
GPT-4o16 / 22
Grok 37 / 22
Grok 3 Mini12 / 22
Grok 49 / 22
Llama 3.1 405B20 / 22
Llama 3.3 70B19 / 22
Llama 4 Maverick17 / 22
Llama 4 Scout22 / 22
Llama-3.1-Nemotron-70B-Instruct18 / 22
MiMo V2.6 Pro14 / 29
Mistral Large8 / 22
Muse Glimmer27 / 29
Muse Spark 1.112 / 31
Muse Spark 1.36 / 29
NVIDIA Nemotron 3 Ultra 550B A55B25 / 29
o35 / 22
Qwen 3.8 27B22 / 29
Qwen3.8 Max16 / 29
Step 516 / 29