Noometry

Knowledge benchmark

GPQA Diamond leaderboard

As of October 2026, GPT-6 Astra has the highest published GPQA Diamond score on Noometry at 95.8%, out of 186 models with results.

Last verified

About GPQA Diamond

198 graduate-level multiple-choice questions in biology, physics and chemistry, written so that skilled non-experts with web access still score poorly.

Category
Knowledge
Introduced
2023
Size
198 questions
Format
Multiple choice (4 options)
Unit
Percent (random guessing ≈ 25%)
Official site
arxiv.org

Top 15 models

  1. GPT-6 Astra 95.8%
  2. Claude Sonnet 5.5 95.6%
  3. Gemini 3.8 Flash 95.4%
  4. GPT-6.1 Sol 95.4%
  5. Gemini 3.7 Flash 94.8%
  6. GPT-5.4 Pro 94.6%
  7. Gemini 3.1 Pro Preview 94.4%
  8. GPT-6 Sol 94.3%
  9. Gemini 3.6 Flash 94.1%
  10. GPT-5.5 94%
  11. Grok 4.6 94%
  12. GPT-5.5 Pro 93.9%
  13. Claude Opus 5 93.9%
  14. GPT-5.6 Sol 93.5%
  15. Grok 4.5 93.4%

Sponsored placements are available on pages like this one. Advertise on Noometry

All results

#ModelProviderScoreSettingSourceDate
1GPT-6 Astra OpenAI95.8%maxEpoch AI2026-08-30
2Claude Sonnet 5.5 Anthropic95.6%maxEpoch AI2026-09-29
3Gemini 3.8 Flash Google95.4%highEpoch AI2026-09-02
4GPT-6.1 Sol OpenAI95.4%maxEpoch AI2026-09-29
5Gemini 3.7 Flash Google94.8%highEpoch AI2026-08-14
6GPT-5.4 Pro OpenAI94.6%xhighEpoch AI2026-03-20
7Gemini 3.1 Pro Preview Google94.4%highEpoch AI2026-08-06
8GPT-6 Sol OpenAI94.3%maxEpoch AI2026-09-22
9Gemini 3.6 Flash Google94.1%highEpoch AI2026-08-02
10GPT-5.5 OpenAI94%xhighEpoch AI2026-04-24
11Grok 4.6 xAI94%highEpoch AI2026-08-12
12GPT-5.5 Pro OpenAI93.9%xhighEpoch AI2026-04-24
13Claude Opus 5 Anthropic93.9%maxEpoch AI2026-07-24
14GPT-5.6 Sol OpenAI93.5%maxEpoch AI2026-07-09
15Grok 4.5 xAI93.4%highEpoch AI2026-07-08
16GPT-5.6 Terra OpenAI93.3%maxEpoch AI2026-07-09
17GPT-5.4 OpenAI93.3%xhighEpoch AI2026-03-06
18Kimi K3 Moonshot AI93.1%maxEpoch AI2026-07-16
19Gemini 3.5 Flash Google92.8%highEpoch AI2026-05-22
20Grok 4.7 xAI92.7%xhighEpoch AI2026-09-22
21Qwen3.8 Max Alibaba (Qwen)92.7%xhighEpoch AI2026-08-04
22Gemini 3 Pro Google92.6%Epoch AI2025-11-19
23GLM-5.2 Z.ai (Zhipu)91.9%maxEpoch AI2026-06-24
24DeepSeek V4 Pro DeepSeek91.7%maxEpoch AI2026-08-18
25GPT-5.6 Luna OpenAI91.6%maxEpoch AI2026-07-09
26GPT-5.2 OpenAI91.4%xhighEpoch AI2025-12-13
27Claude Opus 4.8 Anthropic91%maxEpoch AI2026-06-07
28DeepSeek V4 Flash DeepSeek91%maxEpoch AI2026-08-02
29GLM-5.3 Z.ai (Zhipu)90.9%maxEpoch AI2026-08-24
30MiniMax-M3 MiniMax90.9%Epoch AI2026-08-10
31Qwen3.7 Max Alibaba (Qwen)90.9%Epoch AI2026-08-07
32Kimi K2.6 Moonshot AI90.8%Epoch AI2026-05-01
33Claude Opus 5.5 Anthropic90.6%maxEpoch AI2026-09-22
34Claude Opus 4.6 Anthropic90.5%32KEpoch AI2026-02-06
35Claude Sonnet 5 Anthropic90.5%xhighEpoch AI2026-07-01
36GPT-6 Luna OpenAI90.5%maxEpoch AI2026-09-22
37Claude Opus 4.7 Anthropic90.2%xhighEpoch AI2026-04-17
38GLM-5.3-Flash Z.ai (Zhipu)90.2%maxEpoch AI2026-08-26
39GLM-5.1 Z.ai (Zhipu)89.9%Epoch AI2026-08-10
40DeepSeek V4.1 Flash DeepSeek89.8%maxEpoch AI2026-10-09
41Muse Spark Meta89.8%Epoch AI2026-04-08
42Claude Haiku 5.5 Anthropic89.6%maxEpoch AI2026-10-09
43Gemini 3 Flash Preview Google89.4%highEpoch AI2026-08-06
44Grok 4.20 (Non-Reasoning) xAI89.3%Epoch AI2026-07-13
45Grok 4.3 xAI88.8%highEpoch AI2026-06-17
46Inkling-Small Thinking Machines Lab88.5%xhighEpoch AI2026-08-14
47Qwen3.6 Plus Alibaba (Qwen)88.4%Epoch AI2026-08-07
48Inkling Thinking Machines Lab88.3%xhighEpoch AI2026-08-05
49Kimi K2.7 Code Moonshot AI87.9%Epoch AI2026-08-07
50Qwen3.7 Plus Alibaba (Qwen)87.9%Epoch AI2026-08-07
51GLM-5 Z.ai (Zhipu)87.8%Epoch AI2026-02-12
52GPT-5.1 OpenAI87.6%highEpoch AI2025-11-13
53Kimi K2.5 Moonshot AI87.6%Epoch AI2026-02-02
54Claude Sonnet 4.6 Anthropic87.4%32KEpoch AI2026-02-20
55Qwen3.6 Max Preview Alibaba (Qwen)87.4%Epoch AI2026-08-07
56Grok 4 xAI87%Epoch AI
57GPT-5.4 mini OpenAI86.9%xhighEpoch AI2026-08-07
58Qwen3.5 397B-A17B Alibaba (Qwen)86.4%noneEpoch AI2026-08-07
59GPT-5 OpenAI86.2%highEpoch AI2025-10-29
60Claude Opus 4.5 Anthropic86%32KEpoch AI2025-11-24
61Claude Fable 5 Anthropic85.9%maxEpoch AI2026-08-06
62Qwen3.6 27B Alibaba (Qwen)85.9%Epoch AI2026-08-07
63Nemotron 3 Ultra NVIDIA85.4%Epoch AI2026-08-10
64Gemini 2.5 Pro Google85.3%Epoch AI2025-11-16
65Qwen3.5 Plus Alibaba (Qwen)84.8%Epoch AI2026-08-07
66Qwen3.6 35B-A3B Alibaba (Qwen)84.8%noneEpoch AI2026-08-07
67Kimi K2 Thinking Turbo Moonshot AI84.2%Epoch AI2025-11-11
68Qwen3.5 35B-A3B Alibaba (Qwen)83.5%Epoch AI2026-08-28
69DeepSeek-V3.2-Exp DeepSeek83.4%Epoch AI2025-12-16
70Gemini 3.5 Flash Lite Google83.3%highEpoch AI2026-08-06
71GLM-4.7 Z.ai (Zhipu)83.3%Epoch AI2026-01-29
72Qwen3.6 Flash Alibaba (Qwen)83.3%Epoch AI2026-08-07
73GPT-5.5 Instant OpenAI82.5%Epoch AI2026-08-02
74Claude Sonnet 4.5 Anthropic82.3%59KEpoch AI2025-10-28
75Qwen3.5-Flash Alibaba (Qwen)82.3%Epoch AI2026-08-07
76Qwen3.7 Flash Alibaba (Qwen)82.3%Epoch AI2026-08-07
77Gemini 3.1 Flash Lite Google81.8%highEpoch AI2026-08-06
78o3 OpenAI81.8%highEpoch AI2025-04-16
79Qwen3 235B-A22B Alibaba (Qwen)80.1%Epoch AI2025-12-11
80Claude 3.7 Sonnet Anthropic79.7%64KEpoch AI2025-05-26
81o4-mini OpenAI79.6%highEpoch AI2025-04-16
82Claude Sonnet 4 Anthropic79.2%59KEpoch AI2025-05-26
83Qwen3.5-9B Alibaba (Qwen)79%Epoch AI2026-08-27
84GPT-5.4 nano OpenAI78.5%highEpoch AI2026-04-14
85Claude Opus 4.1 Anthropic77.3%16KEpoch AI2025-08-05
86o3-mini OpenAI77%highEpoch AI2025-02-13
87o1 OpenAI76.8%highEpoch AI2025-02-13
88DeepSeek-R1 DeepSeek76.3%Epoch AI2025-05-29
89Claude Opus 4 Anthropic76.3%16KEpoch AI2025-05-22
90Grok-3 mini xAI76.3%lowEpoch AI2025-04-10
91Gemma 4 31B IT Google75.8%minimalEpoch AI2026-08-06
92gpt-oss-120b OpenAI75.8%highEpoch AI2025-12-11
93Grok 3 xAI75.8%Epoch AI2025-05-26
94GPT-5 Mini OpenAI75%highEpoch AI2025-10-30
95Gemma 4 26B A4B IT Google73.2%minimalEpoch AI2026-08-06
96Qwen3 Max Alibaba (Qwen)72.6%Epoch AI2025-10-06
97Claude Haiku 4.5 Anthropic71.2%32KEpoch AI2025-10-22
98Qwen3-30B-A3B Alibaba (Qwen)70.1%Epoch AI2026-08-30
99GPT-5 Nano OpenAI69.4%highEpoch AI2025-10-30
100GPT-4.5 OpenAI68.7%Epoch AI2025-02-28
101DeepSeek-V3 DeepSeek67.6%Epoch AI2025-04-01
102Llama 4 Maverick Meta67%Epoch AI2025-04-08
103GPT-4.1 OpenAI66.9%Epoch AI2025-04-14
104GPT-4.1 mini OpenAI65.8%Epoch AI2025-04-14
105Qwen3 32B Alibaba (Qwen)65.7%Epoch AI2026-08-28
106Gemini 2.0 Pro Google65.7%Epoch AI2025-02-07
107QwQ Plus Alibaba (Qwen)65.4%Epoch AI2025-04-11
108QwQ-32B Alibaba (Qwen)65.3%Epoch AI2026-08-28
109DeepSeek-R1-Distill-Qwen-32B DeepSeek64.1%Epoch AI2026-08-28
110Gemini 2.0 Flash (Feb 2025) Google64.1%Epoch AI2025-02-06
111Qwen3 14B Alibaba (Qwen)63.8%Epoch AI2026-08-28
112o1-mini OpenAI62.4%highEpoch AI2025-02-13
113gpt-oss-20b OpenAI60.8%mediumEpoch AI2026-08-27
114GLM-4.7-Flash Z.ai (Zhipu)60.5%Epoch AI2026-08-30
115Mistral Medium Mistral AI59.5%Epoch AI2025-05-07
116Gemini 1.5 Pro (May 2024) Google57.2%Epoch AI2025-01-27
117Qwen3 8B Alibaba (Qwen)56.8%Epoch AI2026-08-27
118Qwen Max Alibaba (Qwen)56.1%Epoch AI2025-04-01
119Magistral Small Mistral AI56.1%Epoch AI2026-08-06
120Phi-4 Microsoft56.1%Epoch AI2025-01-31
121DeepSeek-R1-Distill-Llama-70B DeepSeek55.7%Epoch AI2025-03-10
122Claude 3.5 Sonnet Anthropic55.3%Epoch AI2025-01-27
123Grok-2 (Dec 2024) xAI53.8%Epoch AI2025-01-27
124Qwen3-4B Alibaba (Qwen)52.3%Epoch AI2026-08-28
125Llama 4 Scout Meta51.8%Epoch AI2025-04-08
126Mistral Large Mistral AI51.3%Epoch AI2025-02-25
127Llama 3.1-405B Meta50.9%Epoch AI2025-01-27
128GPT-4o OpenAI49.2%Epoch AI2025-01-27
129Qwen2.5 72B Instruct Alibaba (Qwen)49.1%Epoch AI2025-01-27
130Mistral Small 3.2 Mistral AI49.1%Epoch AI2026-08-28
131GPT-4.1 nano OpenAI48.9%Epoch AI2025-04-14
132Qwen Plus Alibaba (Qwen)48.1%Epoch AI2025-04-07
133Gemma 3 27B Google47.7%Epoch AI2026-08-27
134Mistral Small Mistral AI47.5%Epoch AI2025-03-18
135Llama-3.3-70B-Instruct Meta47.4%Epoch AI2025-01-27
136Gemini 1.5 Flash (May 2024) Google47.3%Epoch AI2025-01-27
137Mistral Small 3 Mistral AI47.3%Epoch AI2026-08-30
138Claude 3 Opus Anthropic47.2%Epoch AI2025-01-27
139GPT-4 Turbo OpenAI46.6%Epoch AI2025-02-27
140Tulu 3 (Tülu 3) 70B Allen Institute for AI (Ai2)46.3%Epoch AI2025-01-27
141Qwen2.5 32B Instruct Alibaba (Qwen)46.1%Epoch AI2025-01-30
142DeepSeek-R1-Distill-Qwen-14B DeepSeek44.7%Epoch AI2025-03-10
143Llama 3.1-70B Meta44.2%Epoch AI2025-01-27
144WizardLM-2 8x22B Microsoft43.4%Epoch AI2025-01-27
145Mistral Small 3.1 Mistral AI41.9%Epoch AI2026-08-30
146Qwen Turbo Alibaba (Qwen)41.8%Epoch AI2025-04-07
147Llama 3.2 90B Meta41%Epoch AI2025-01-27
148Qwen2-72B Alibaba (Qwen)40.8%Epoch AI2025-01-27
149Claude 3 Sonnet Anthropic40.6%Epoch AI2025-01-27
150Llama 3-70B Meta40.6%Epoch AI2025-01-27
151Gemma 3 12B Google39.5%Epoch AI2026-08-28
152Claude 3.5 Haiku Anthropic38.1%Epoch AI2025-03-12
153Qwen3-1.7B Alibaba (Qwen)38%Epoch AI2026-08-28
154GPT-4o mini OpenAI37.7%Epoch AI2025-01-27
155Gemma 2 27B Google36.5%Epoch AI2025-01-27
156Claude 3 Haiku Anthropic36.3%Epoch AI2025-01-27
157GPT-4 OpenAI35.7%Epoch AI2025-10-23
158Qwen2.5 7B Instruct Alibaba (Qwen)35.5%Epoch AI2026-08-30
159Claude 2 Anthropic34.7%Epoch AI2025-01-27
160Mixtral 8x22B Mistral AI34.1%Epoch AI2025-01-27
161Gemini 1.0 Pro Google34%Epoch AI2025-01-27
162Eurus-2-7B-PRIME Shanghai AI Lab33.9%Epoch AI2025-01-27
163DeepSeek-R1-Distill-Qwen-1.5B DeepSeek33.6%Epoch AI2026-08-28
164Claude 2.1 Anthropic33%Epoch AI2025-01-27
165Gemini 1.5 Flash 8B Google33%Epoch AI2025-03-12
166DBRX Databricks32.9%Epoch AI2025-01-27
167Yi-1.5-34B 01.AI32%Epoch AI2025-01-27
168Qwen1.5-32B Alibaba (Qwen)30.7%Epoch AI2025-01-27
169Mixtral 8x7B Mistral AI30.6%Epoch AI2025-01-27
170Mistral Nemo Mistral AI29.9%Epoch AI2025-01-27
171Qwen1.5-72B Alibaba (Qwen)28.8%Epoch AI2025-01-27
172Granite 4.0 Micro IBM28.3%Epoch AI2026-08-30
173GPT-3.5-turbo OpenAI28%Epoch AI2025-01-27
174phi-3-medium 14B Microsoft27.6%Epoch AI2025-01-31
175Gemma 2 9B Google27.5%Epoch AI2025-01-27
176Ministral 8B Mistral AI27.1%Epoch AI2025-01-27
177Llama 3.1-8B Meta27%Epoch AI2026-08-27
178Llama 2-70B Meta26.3%Epoch AI2025-01-27
179Llama 3-8B Meta26.1%Epoch AI2025-01-27
180Ministral 3B Mistral AI25.3%Epoch AI2025-01-27
181DeepSeek LLM 67B DeepSeek24.6%Epoch AI2025-01-27
182Llama 3.2 1B Meta23.9%Epoch AI2026-08-30
183Gemma 3 4B Google23.2%Epoch AI2026-08-28
184Gemma 3 1B Google19.9%Epoch AI2026-08-28
185Mistral 7B Mistral AI15.2%Epoch AI2025-01-27
186Yi-34B 01.AI14.7%Epoch AI2025-01-27

Compare the leaders

Other knowledge benchmarks

Frequently asked questions

What does GPQA Diamond measure?

198 graduate-level multiple-choice questions in biology, physics and chemistry, written so that skilled non-experts with web access still score poorly.

Which model has the highest GPQA Diamond score?

As of October 2026, GPT-6 Astra has the highest published GPQA Diamond score on Noometry at 95.8%, out of 186 models with results.

What is the best open-weight model on GPQA Diamond?

Kimi K3 has the highest GPQA Diamond accuracy among open-weight models at 93.1%, ranking 18 of 186 overall.