Artificial Analysis GPQA Diamond

An Artificial Analysis GPQA Diamond score.

ModelsModels with a result.180
Top resultBest result on this test.96.1%GPT-6 Astra
Top-3 spreadPoints from first to third.1.2 pts
YearYear of release.2026

Result and price

020406080100$0.03$0.1$0.3$1$3$10$30$100RESULTOUTPUT PRICE PER 1M TOKENS ยท LOG SCALE

Results

180 results
#ModelResultThe score from the source.
03Grok 4.694.9%
08GPT-5.593.5%
08Kimi K393.5%
12Grok 4.593.1%
13MiniMax M392.9%
22GPT-5.492.0%
24GLM-5.391.7%
35GPT-5.290.3%
36Grok 4.390.1%
40Hy389.7%
44GLM-5.289.5%
49Muse Spark88.4%
53Grok 487.7%
56GPT-5.187.3%
57Inkling87.2%
59GLM-5.186.8%
70GLM-4.785.9%
72A.X K285.7%
81o3-pro84.5%
92o382.7%
93GLM-582.0%
104K-Exaone78.3%
105GPT-OSS 120B78.2%
108Kimi K276.6%
109o1-preview76.5%
110Qwen3 Max76.4%
111Command A+76.1%
114Gemma 4 12B75.3%
118o3-mini74.8%
120o174.7%
125GLM-4.5-Air73.3%
132GPT-OSS 20B68.8%
137GPT-4.166.6%
138GPT-4.1 mini66.4%
142Sarvam 30B63.3%
143GLM-4.663.2%
144Celeris-163.1%
153Gemma 4 E4B57.6%
154Phi-457.5%
158DeepSeek V355.7%
159GPT-4o54.3%
162GPT-4.1 nano51.2%
163Nova Pro49.9%
168Gemma 4 E2B43.3%
169Gemma 3 27B42.8%
170GPT-4o mini42.6%