benchgap
Knowledge & reasoning

HealthBench Professional (raw, Artificial Analysis) leaderboard

As of 2026-11-18, the highest measured score on HealthBench Professional (raw, Artificial Analysis) is 71.4% by Claude Opus 5.5. 205 more models have estimated scores, calibrated from the benchmarks they were measured on.

Measured scores: benchlm.ai.

#ModelScoreSource
1Claude Opus 5.571.4%measured
2GPT-6.2 Astra71.4%estimated ± 1.1 pp, medium confidence
3Claude Fable 5.271.1%measured
4Gemini 4 Carbon71.0%estimated ± 1.1 pp, medium confidence
5Claude Sonnet 5.570.8%measured
6Muse Wonder Full70.7%estimated ± 0.8 pp, medium confidence
7GPT-5.6 Sol70.6%estimated ± 0.8 pp, medium confidence
8GPT-6.2 Sol70.5%estimated ± 1.1 pp, medium confidence
9GPT-6.1 Sol70.4%estimated ± 0.8 pp, medium confidence
10GPT-6 Sol70.2%estimated ± 0.8 pp, medium confidence
11GPT-5.5 Pro70.1%estimated ± 0.8 pp, medium confidence
12Gemini 4 Hydrogen69.9%estimated ± 1.1 pp, medium confidence
13GPT-5.4 Pro69.9%estimated ± 0.8 pp, medium confidence
14GPT-5.6 Terra69.9%estimated ± 0.8 pp, medium confidence
15GPT-6 Astra69.7%measured
16Claude Opus 569.6%estimated ± 0.8 pp, medium confidence
17Claude Mythos 569.4%estimated ± 2.6 pp, medium confidence
18Claude Fable 569.4%estimated ± 0.8 pp, medium confidence
19Muse Wonder69.3%estimated ± 0.8 pp, medium confidence
20GPT-6.1 Terra69.2%estimated ± 0.8 pp, medium confidence
21Claude Fable 5.169.2%measured
22Gemini 4 Argon68.9%estimated ± 0.8 pp, medium confidence
23GPT-5.568.9%estimated ± 0.8 pp, medium confidence
24MiMo-V2.6-Pro68.7%estimated ± 0.8 pp, medium confidence
25Muse Spark 1.368.0%estimated ± 0.8 pp, medium confidence
26Kimi K3.1 Pro67.9%estimated ± 0.8 pp, medium confidence
27GPT-5.467.3%estimated ± 0.8 pp, medium confidence
28Kimi K367.3%estimated ± 0.8 pp, medium confidence
29Qwen4 Max67.2%estimated ± 1.1 pp, medium confidence
30Ling 3.0 Tiny67.1%estimated ± 2.5 pp, medium confidence
31Sakana Fugu-Ultra67.1%estimated ± 2.6 pp, medium confidence
32MiniMax M3.1 Flash66.8%estimated ± 0.8 pp, medium confidence
33Pareto 26.966.8%estimated ± 2.6 pp, medium confidence
34Sakana Fugu66.2%estimated ± 2.6 pp, medium confidence
35Claude Opus 4.866.0%estimated ± 0.8 pp, medium confidence
36GLM-5.266.0%estimated ± 0.8 pp, medium confidence
37Step 5 Preview66.0%estimated ± 0.8 pp, medium confidence
38ERNIE 5.5 Flash Preview65.9%estimated ± 2.6 pp, medium confidence
39GPT-5.6 Luna65.9%estimated ± 0.8 pp, medium confidence
40Sage-A1.565.5%estimated ± 0.8 pp, medium confidence
41Ornith-1.5-397B65.3%estimated ± 2.6 pp, low confidence
42GPT-6 Luna65.1%estimated ± 0.8 pp, medium confidence
43Qwen3.8 Max64.9%estimated ± 2.6 pp, low confidence
44GLM-5.364.9%estimated ± 0.8 pp, medium confidence
45Muse Ember64.9%estimated ± 1.1 pp, low confidence
46Claude Haiku 5.564.8%measured
47Gemini 3.8 Flash64.4%estimated ± 0.8 pp, low confidence
48DeepSeek V4 Pro64.2%estimated ± 0.8 pp, low confidence
49Ling 3.1 Flash64.2%estimated ± 0.8 pp, low confidence
50Gemini 3.1 Pro63.9%estimated ± 0.8 pp, low confidence
51Grok 4.763.9%estimated ± 0.8 pp, low confidence
52Muse Spark 1.263.9%estimated ± 0.8 pp, low confidence
53Qwen3.8 Max Preview63.9%estimated ± 0.8 pp, low confidence
54Beam 1.0 501B-A23B FP863.7%estimated ± 0.8 pp, low confidence
55Grok 4.663.5%estimated ± 0.8 pp, low confidence
56Claude Sonnet 563.3%estimated ± 0.8 pp, low confidence
57GPT-5.3 Codex63.3%estimated ± 0.8 pp, low confidence
58Hy4 preview63.3%estimated ± 0.8 pp, low confidence
59DeepSeek V4 Flash 073163.1%estimated ± 0.8 pp, low confidence
60Beam 1.0 501B-A23B62.8%estimated ± 0.8 pp, low confidence
61Inkling 2 Small62.6%estimated ± 1.1 pp, low confidence
62Grok 4.562.0%estimated ± 0.8 pp, low confidence
63Beam 1.0 501B-A23B NVFP462.0%estimated ± 0.8 pp, low confidence
64Muse Spark 1.161.7%estimated ± 0.8 pp, low confidence
65Muse Spark Open61.5%estimated ± 0.8 pp, low confidence
66MiMo-V2-Pro61.5%estimated ± 1.1 pp, low confidence
67GLM-561.1%estimated ± 1.1 pp, low confidence
68DeepSeek V4.1 Flash 101760.9%estimated ± 0.8 pp, low confidence
69Gemini 3.7 Flash60.9%estimated ± 0.8 pp, low confidence
70GLM-5-Turbo60.3%estimated ± 1.1 pp, low confidence
71Qwen3.7 Max59.9%estimated ± 0.8 pp, low confidence
72Gemini 3.5 Flash59.6%estimated ± 0.8 pp, low confidence
73Grok 4.2059.3%estimated ± 2.6 pp, low confidence
74Claude Sonnet 4.659.1%estimated ± 1.1 pp, low confidence
75Claude Opus 4.6 (Adaptive)59.0%estimated ± 0.8 pp, low confidence
76MiMo-V2-Omni58.6%estimated ± 1.1 pp, low confidence
77Claude Opus 4.558.4%estimated ± 1.1 pp, low confidence
78GLM-5V-Turbo58.3%estimated ± 1.1 pp, low confidence
79Claude Opus 4.758.2%estimated ± 0.8 pp, low confidence
80MiMo-V2.6-Flash58.2%estimated ± 0.8 pp, low confidence
81Qwen3.5-27B57.8%estimated ± 1.1 pp, low confidence
82MiniMax M2.757.7%estimated ± 1.1 pp, low confidence
83GPT-5.257.6%estimated ± 0.8 pp, low confidence
84Solar Open 257.6%estimated ± 2.6 pp, low confidence
85Command A+57.5%estimated ± 1.1 pp, low confidence
86Grok 457.5%estimated ± 1.1 pp, low confidence
87GLM-4.757.3%estimated ± 1.1 pp, low confidence
88Muse Spark57.2%estimated ± 0.8 pp, low confidence
89o3-pro57.0%estimated ± 1.1 pp, low confidence
90Qwen3.8-Flash-Next56.9%estimated ± 0.8 pp, low confidence
91Qwen3.6-27B56.6%estimated ± 1.1 pp, low confidence
92Nemotron 3.5 Super VL 120B A12B56.5%estimated ± 0.8 pp, low confidence
93Gemini 3.6 Flash56.1%estimated ± 0.8 pp, low confidence
94Nemotron 3.5 Super VL 120B A12B NVFP456.1%estimated ± 0.8 pp, low confidence
95Claude 4 Opus Thinking56.0%estimated ± 1.1 pp, low confidence
96Ling 3.0 Flash55.5%estimated ± 1.1 pp, low confidence
97Ornith-1.5-35B-A3B55.3%estimated ± 2.6 pp, low confidence
98GPT-5.4 mini55.1%estimated ± 0.8 pp, low confidence
99Kimi K2.7 Code55.1%estimated ± 0.8 pp, low confidence
100Mistral Large 455.1%estimated ± 0.8 pp, low confidence
101Step 3.7 Flash54.9%estimated ± 1.1 pp, low confidence
102Qwen3.5-35B-A3B54.8%estimated ± 1.1 pp, low confidence
103Quasar 438B54.0%estimated ± 0.8 pp, low confidence
104Claude 4.1 Opus54.0%estimated ± 1.1 pp, low confidence
105GPT-5.4 nano53.8%estimated ± 0.8 pp, low confidence
106Gemini 3 Pro53.4%estimated ± 0.8 pp, low confidence
107Qwen3.7 Plus53.4%estimated ± 0.8 pp, low confidence
108Gemini 3 Flash53.4%estimated ± 1.1 pp, low confidence
109Claude 3.7 Sonnet Thinking53.1%estimated ± 1.1 pp, low confidence
110A.X K253.0%estimated ± 0.8 pp, low confidence
111GPT-5.2-Codex52.6%estimated ± 0.8 pp, low confidence
112o4-mini51.9%estimated ± 1.1 pp, low confidence
113Claude 4 Opus51.9%estimated ± 1.1 pp, low confidence
114Claude 4 Sonnet51.9%estimated ± 1.1 pp, low confidence
115Inkling-Small51.8%estimated ± 0.8 pp, low confidence
116MiMo-V2-Flash51.2%estimated ± 1.1 pp, low confidence
117DeepSeek V3.251.2%estimated ± 1.1 pp, low confidence
118Grok 4.351.1%estimated ± 0.8 pp, low confidence
119Kimi K2.651.1%estimated ± 0.8 pp, low confidence
120Qwen3 Max50.7%estimated ± 1.1 pp, low confidence
121Ornith-1.5-9B50.4%estimated ± 2.6 pp, low confidence
122Claude 3.7 Sonnet50.3%estimated ± 1.1 pp, low confidence
123o150.2%estimated ± 1.1 pp, low confidence
124GLM-4.649.8%estimated ± 1.1 pp, low confidence
125K-Exaone49.1%estimated ± 1.1 pp, low confidence
126Grok Code Fast 148.6%estimated ± 1.1 pp, low confidence
127Ling 2.6 Flash48.6%estimated ± 1.1 pp, low confidence
128DeepSeek V3.148.1%estimated ± 1.1 pp, low confidence
129DeepSeek-R147.2%estimated ± 1.1 pp, low confidence
130Kimi K246.6%estimated ± 1.1 pp, low confidence
131GPT-4.146.5%estimated ± 1.1 pp, low confidence
132Solar Pro 446.5%estimated ± 0.8 pp, low confidence
133o3-mini46.2%estimated ± 1.1 pp, low confidence
134o1-pro46.1%estimated ± 1.1 pp, low confidence
135Gemini 3.5 Flash-Lite44.8%estimated ± 1.1 pp, low confidence
136GPT-544.5%estimated ± 0.8 pp, low confidence
137GPT-5.1-Codex44.5%estimated ± 0.8 pp, low confidence
138GPT-5.1-Codex-Max44.5%estimated ± 0.8 pp, low confidence
139o1-preview44.3%estimated ± 1.1 pp, low confidence
140Grok 4.1 Fast44.1%estimated ± 1.1 pp, low confidence
141GLM-4.5-Air43.8%estimated ± 1.1 pp, low confidence
142Inkling43.4%estimated ± 0.8 pp, low confidence
143Qwen3.8-27B43.4%estimated ± 0.8 pp, low confidence
144GPT-4.1 mini42.0%estimated ± 1.1 pp, low confidence
145GPT-5.141.4%estimated ± 0.8 pp, low confidence
146Hy341.4%estimated ± 0.8 pp, low confidence
147Hy3 Preview41.4%estimated ± 0.8 pp, low confidence
148Gemini 2.5 Flash41.3%estimated ± 1.1 pp, low confidence
149o1-mini41.2%estimated ± 1.1 pp, low confidence
150DeepSeek V3 032441.1%estimated ± 1.1 pp, low confidence
151GPT-4.540.8%estimated ± 1.1 pp, low confidence
152QwQ-32B40.5%estimated ± 1.1 pp, low confidence
153Apodex 1.140.2%estimated ± 0.8 pp, low confidence
154Apodex 1.1 Mini40.2%estimated ± 0.8 pp, low confidence
155Claude Opus 4.5 Thinking40.2%estimated ± 0.8 pp, low confidence
156GLM-5.140.2%estimated ± 0.8 pp, low confidence
157Mistral Large 340.1%estimated ± 1.1 pp, low confidence
158Magistral Medium39.6%estimated ± 1.1 pp, low confidence
159Mistral Medium 339.6%estimated ± 1.1 pp, low confidence
160Claude 3.5 Haiku39.3%estimated ± 1.1 pp, low confidence
161Gemini 2.0 Flash39.3%estimated ± 1.1 pp, low confidence
162Claude 3 Opus38.8%estimated ± 1.1 pp, low confidence
163Nemotron 3.5 Bolt 74B-A9B38.8%estimated ± 0.8 pp, low confidence
164DeepSeek V338.3%estimated ± 1.1 pp, low confidence
165GPT-4o38.2%estimated ± 1.1 pp, low confidence
166MiMo-V2.5-Pro37.4%estimated ± 0.8 pp, low confidence
167MiniMax M335.8%estimated ± 0.8 pp, low confidence
168Qwen 3.6 Max (preview)35.8%estimated ± 0.8 pp, low confidence
169Mistral Medium 3.5 128B34.4%estimated ± 2.5 pp, low confidence
170Kimi K2.532.3%estimated ± 0.8 pp, low confidence
171Nemotron 3 Ultra32.3%estimated ± 0.8 pp, low confidence
172Grok 4.1 Fast (Reasoning)31.0%estimated ± 0.8 pp, low confidence
173Grok 4 Fast (Reasoning)31.0%estimated ± 0.8 pp, low confidence
174Qwen3.6 Plus31.0%estimated ± 0.8 pp, low confidence
175Qwen3.5 397B30.6%estimated ± 2.1 pp, low confidence
176Claude Opus 4.630.4%estimated ± 0.8 pp, low confidence
177Mercury 2.529.7%estimated ± 2.5 pp, low confidence
178Gemini 2.5 Pro29.0%estimated ± 0.8 pp, low confidence
179Muse Glimmer 30B29.0%estimated ± 0.8 pp, low confidence
180Nemotron 3.5 Bolt 74B-A9B BF1628.8%estimated ± 2.5 pp, low confidence
181Mistral Small 426.9%estimated ± 2.5 pp, low confidence
182Mistral Large 220.1%estimated ± 2.5 pp, low confidence
183Qwen3.5-122B-A10B17.4%estimated ± 2.1 pp, low confidence
184Solar Pro 314.5%estimated ± 2.5 pp, low confidence
185North Mini Code14.1%estimated ± 2.5 pp, low confidence
186Gemma 4 31B12.7%estimated ± 2.1 pp, low confidence
187Magistral Small12.7%estimated ± 2.5 pp, low confidence
188Claude 3 Haiku11.3%estimated ± 2.5 pp, low confidence
189Qwen3.6-35B-A3B9.3%estimated ± 2.1 pp, low confidence
190Nova Pro9.2%estimated ± 2.5 pp, low confidence
191Ling 3.0 Flash VL8.9%estimated ± 2.1 pp, low confidence
192Solar Pro 27.1%estimated ± 2.5 pp, low confidence
193Gemini 2.5 Flash-Lite5.6%estimated ± 2.5 pp, low confidence
194o35.5%estimated ± 2.1 pp, low confidence
195Mistral Small 3.15.0%estimated ± 2.5 pp, low confidence
196GPT-OSS 120B4.7%estimated ± 2.1 pp, low confidence
197Gemma 4 26B A4B4.3%estimated ± 2.1 pp, low confidence
198Devstral Small4.3%estimated ± 2.5 pp, low confidence
199Nemotron 3.5 Lightning 30B A3B NVFP44.3%estimated ± 2.5 pp, low confidence
200K-EXAONE 2.03.5%estimated ± 2.1 pp, low confidence
201GPT-4.1 nano2.9%estimated ± 2.5 pp, low confidence
202Gemma 3 27B2.1%estimated ± 2.5 pp, low confidence
203Trinity-Large-Preview1.4%estimated ± 2.1 pp, low confidence
204Trinity-Large-Thinking1.4%estimated ± 2.1 pp, low confidence
205Gemma 4 12B1.3%estimated ± 2.1 pp, low confidence
206Granite 4.2 8B1.2%estimated ± 2.5 pp, low confidence
207Granite 4.2 30B0.9%estimated ± 2.5 pp, low confidence
208Gemma 4 E4B0.4%estimated ± 2.5 pp, low confidence
209LFM2.5-8B-A1B0.3%estimated ± 2.5 pp, low confidence
210Granite 4.2 3B0.3%estimated ± 2.5 pp, low confidence
211LFM3-1.7B0.1%estimated ± 2.5 pp, low confidence

All leaderboards

Agentic · terminal

Agentic · tools

Coding

Math

Knowledge & reasoning

Instruction following

Multilingual

Vision & documents

Long context

General