Benchmarks / Artificial Analysis

Reported by Artificial Analysis

Artificial Analysis

Share of expert-written exam questions answered correctly, as run by Artificial Analysis with its own harness and prompts.

Last updated 9 Nov 2026

Results dated
20 Oct 2026 to 9 Nov 2026
Results
419 configurations of 250 models
Unit
% of questions
Licence
Artificial Analysis commercial data licence

Artificial Analysis in brief

What it measures. Artificial Analysis is a third-party benchmark, shown here with credit. Measured here: Share of expert-written exam questions answered correctly, as run by Artificial Analysis with its own harness and prompts.

Latest results. , 250 models ranked.

Top 3 on Humanity's Last Exam (% of questions, higher is better, each model at its best setting):

  1. Claude Opus 5.5 (max reasoning) (Anthropic): 61.4%
  2. Claude Fable 5.1 (max reasoning) (Anthropic): 59.1%
  3. Gemini 4 Argon (high reasoning) (Google): 57.1%

How to cite. Artificial Analysis, via Spring Prompt, results dated 9 November 2026. https://springprompt.com/benchmarks/artificial-analysis These results belong to Artificial Analysis, not to us. Licence: Artificial Analysis commercial data licence. Credit Artificial Analysis first.

Humanity's Last Exam: Kimi K2.5

Top 15 of 250 results · % of questions, higher is better. Choose a model to highlight it.Clear highlight

  1. 1 Claude Opus 5.5 (max reasoning)Anthropic 61.4%
  2. 2 Claude Fable 5.1 (max reasoning)Anthropic 59.1%
  3. 3 Gemini 4 Argon (high reasoning)Google 57.1%
  4. 4 Claude Fable 5 (max reasoning)Anthropic 55.5%
  5. 5 Claude Sonnet 5.5 (max reasoning)Anthropic 55.0%
  6. 6 Claude Opus 5 (max reasoning)Anthropic 54.9%
  7. 7 GPT-6 Astra (max reasoning)OpenAI 54.7%
  8. 8 GPT-6.1 Sol (max reasoning)OpenAI 52.9%
  9. 9 Kimi K3.1 ProMoonshot AI 52.5%
  10. 10 Muse Wonder (max reasoning)Meta 50.1%
  11. 11 GPT-5.6 Sol (max reasoning)OpenAI 49.5%
  12. 12 MiMo-V2.6-ProXiaomi 49.4%
  13. 13 Claude Opus 4.8 (max reasoning)Anthropic 48.7%
  14. 13 Muse Spark 1.3 (max reasoning)Meta 48.7%
  15. 15 MiniMax M3.1 FlashMiniMax 48.0%
  16. 78 Kimi K2.5Moonshot AI 30.7%

Full results

Artificial Analysis: Humanity's Last Exam, % of questions, higher is better
#ModelHumanity's Last Exam
% of questions, higher is better
Price
$ per million tokens, in / out
1 Claude Opus 5.5 (max reasoning)Anthropic · best of 5 settings
61.4%
$4 / $20
2 Claude Fable 5.1 (max reasoning)Anthropic · best of 5 settings
59.1%
$10 / $50
3 Gemini 4 Argon (high reasoning)Google
57.1%
$2 / $10
4 Claude Fable 5 (max reasoning)Anthropic
55.5%
$10 / $50
5 Claude Sonnet 5.5 (max reasoning)Anthropic · best of 5 settings
55.0%
$2 / $10
6 Claude Opus 5 (max reasoning)Anthropic · best of 5 settings
54.9%
$5 / $25
7 GPT-6 Astra (max reasoning)OpenAI · best of 5 settings
54.7%
$10 / $50
8 GPT-6.1 Sol (max reasoning)OpenAI · best of 5 settings
52.9%
$2 / $10
9 Kimi K3.1 ProMoonshot AI
52.5%
$1.75 / $10
10 Muse Wonder (max reasoning)Meta
50.1%
$4 / $20
11 GPT-5.6 Sol (max reasoning)OpenAI · best of 6 settings
49.5%
$4 / $20
12 MiMo-V2.6-ProXiaomi
49.4%
$0.43 / $0.87
13 Claude Opus 4.8 (max reasoning)Anthropic
48.7%
$5 / $25
13 Muse Spark 1.3 (max reasoning)Meta · best of 2 settings
48.7%
$1.25 / $4.25
15 MiniMax M3.1 FlashMiniMax
48.0%
$0.15 / $0.60
16 Gemini 3.7 Flash (high reasoning)Google · best of 3 settings
47.9%
$1.50 / $7.50
16 GPT-6 Sol (max reasoning)OpenAI · best of 6 settings
47.9%
$2 / $10
18 Gemini 3.8 Flash (high reasoning)Google · best of 3 settings
47.8%
$1.50 / $7.50
19 Gemini 3.1 Pro PreviewGoogle
47.0%
$2 / $12
20 Kimi K3 (max reasoning)Moonshot AI · best of 2 settings
46.9%
$3 / $15
21 Muse Spark 1.1 (extra-high reasoning)Meta
46.2%
$1.25 / $4.25
21 GPT-6.1 Terra (max reasoning)OpenAI · best of 6 settings
46.2%
$1 / $5
23 GPT-5.5 (extra-high reasoning)OpenAI · best of 5 settings
45.8%
$5 / $30
24 Muse Spark 1.2 (extra-high reasoning)Meta
45.5%
$1.25 / $4.25
25 Claude Haiku 5.5 (max reasoning)Anthropic · best of 5 settings
44.4%
$0.10 / $0.50
26 Grok 4.6 (extra-high reasoning)xAI · best of 4 settings
44.1%
$2 / $6
27 Muse Spark Open (extra-high reasoning)Meta
43.9%
$1.25 / $4.25
28 GPT-5.4 (extra-high reasoning)OpenAI · best of 3 settings
43.7%
$2.50 / $15
29 DeepSeek-V4.1-Flash-1017 (max reasoning)DeepSeek · best of 2 settings
43.3%
$0.30 / $1.20
30 Qwen3.8-Max (0902)Alibaba
43.1%
$2 / $6
30 Grok 4.7 (extra-high reasoning)xAI · best of 3 settings
43.1%
$2 / $6
32 Qwen3.8-Max (0803)Alibaba
43.0%
–
33 GPT-5.6 Terra (max reasoning)OpenAI · best of 6 settings
42.9%
$2 / $12
34 Gemini 3.5 Flash (high reasoning)Google · best of 3 settings
42.7%
$1.50 / $9
34 Grok 4.5 (high reasoning)xAI
42.7%
$2 / $6
36 GPT-5.3-Codex (extra-high reasoning)OpenAI
42.5%
$1.75 / $14
37 Qwen3.8-2.4T-A95BAlibaba
42.4%
$2 / $6
38 Claude Opus 4.7 (max reasoning)Anthropic · best of 2 settings
42.3%
$5 / $25
38 GLM-5.3 (max reasoning)Z.ai · best of 2 settings
42.3%
$1.40 / $4.40
40 Claude Sonnet 5 (max reasoning)Anthropic · best of 6 settings
41.3%
$2 / $10
41 GLM-5.2 (max reasoning)Z.ai · best of 2 settings
41.1%
$1.40 / $4.40
42 DeepSeek-V4-Pro (0813, max reasoning)DeepSeek · best of 2 settings
41.0%
$1.32 / $3.96
43 Gemini 3.6 Flash (high reasoning)Google
40.8%
$1.50 / $7.50
44 Muse SparkMeta
40.7%
–
45 Qwen3.7-MaxAlibaba
40.5%
$1.48 / $4.42
46 Claude Opus 4.6 (max reasoning)Anthropic · best of 2 settings
39.9%
$5 / $25
46 GLM-5.3-FlashZ.ai
39.9%
$0.15 / $0.50
48 Gemini 3 Pro Preview (high reasoning)Google · best of 2 settings
39.7%
–
49 GPT-5.6 Luna (max reasoning)OpenAI · best of 6 settings
39.5%
$0.20 / $1.20
50 DeepSeek-V4.1-Flash (max reasoning)DeepSeek · best of 2 settings
39.2%
$0.30 / $1.20
51 MiniMax-M3MiniMax
39.0%
$0.30 / $1.20
52 DeepSeek-V4-Flash (0731, max reasoning)DeepSeek
38.6%
$0.13 / $0.40
53 GPT-6 Luna (max reasoning)OpenAI · best of 6 settings
38.5%
$0.10 / $0.50
54 Grok Build 0.1xAI
38.3%
$1 / $2
55 Qwen3.8-Flash-NextAlibaba
38.0%
–
56 GPT-5.2 (extra-high reasoning)OpenAI · best of 3 settings
37.7%
$1.75 / $14
57 DeepSeek-V4-Pro (0423, max reasoning)DeepSeek · best of 3 settings
37.5%
$1.42 / $2.83
57 Kimi K2.6Moonshot AI · best of 2 settings
37.5%
$0.95 / $4
59 Grok 4.3 (high reasoning)xAI · best of 4 settings
37.2%
$1.25 / $2.50
60 Gemini 3 Flash Preview (reasoning on)Google · best of 2 settings
36.6%
$0.50 / $3
61 GPT-5.2-Codex (extra-high reasoning)OpenAI
35.7%
$1.75 / $14
61 MiMo-V2.5-Pro (reasoning on)Xiaomi · best of 2 settings
35.7%
$0.43 / $0.87
63 Qwen3.7-PlusAlibaba
35.6%
$0.32 / $1.28
64 MiMo-V2.6-FlashXiaomi
35.1%
$0.14 / $0.28
65 Mistral Large 4Mistral AI
35.0%
$1.36 / $4.18
65 Kimi K2.7 CodeMoonshot AI
35.0%
$0.95 / $4
67 DeepSeek-V4-Flash (0423, max reasoning)DeepSeek · best of 3 settings
34.8%
$0.05 / $0.45
68 Beam 1.0 501B A23BReflection
34.7%
$0.40 / $1.20
69 DeepSeek-V4-Flash-Vision-Exp (max reasoning)DeepSeek
34.5%
$0.22 / $0.65
69 Grok 4.20xAI · best of 2 settings
34.5%
$1.25 / $2.50
71 Qwen3.8-27B (extra-high reasoning)Alibaba · best of 4 settings
33.9%
$0.50 / $3
72 Claude Sonnet 4.6 (max reasoning)Anthropic · best of 2 settings
33.6%
$3 / $15
73 Hy3Tencent
33.5%
$0.14 / $0.58
74 Inkling SmallThinking Machines
33.3%
$0.45 / $1.20
75 Grok 4.20 (0309, reasoning)xAI
32.4%
–
76 Inkling (extra-high reasoning)Thinking Machines
31.9%
$1 / $4.05
77 Qwen3.6-Max-PreviewAlibaba
30.8%
–
78 Kimi K2.5Moonshot AI · best of 2 settings
30.7%
$0.57 / $2.85
78 Nemotron 3.5 Super VLNVIDIA
30.7%
$0.15 / $0.50
80 Claude Opus 4.5 (reasoning on)Anthropic · best of 2 settings
30.1%
$5 / $25
80 GLM-5.1Z.ai · best of 2 settings
30.1%
$1.33 / $4.18
82 Solar Pro 4Upstage
29.8%
$0.15 / $0.60
83 MiniMax-M2.7MiniMax
29.6%
$0.30 / $1.20
84 GLM-5Z.ai · best of 2 settings
29.3%
$1 / $3.20
85 Qwen3.5-397B-A17BAlibaba · best of 2 settings
29.0%
$0.55 / $3.50
86 DeepSeek-V3.2-SpecialeDeepSeek
28.7%
–
87 GPT-5.1 (high reasoning)OpenAI · best of 2 settings
28.5%
$1.25 / $10
87 GPT-5 (high reasoning)OpenAI · best of 4 settings
28.5%
$1.25 / $10
89 GPT-5.4 nano (extra-high reasoning)OpenAI · best of 3 settings
28.3%
$0.20 / $1.25
90 GPT-5.4 mini (extra-high reasoning)OpenAI · best of 3 settings
28.1%
$0.75 / $4.50
91 Qwen3-Max-ThinkingAlibaba
28.0%
–
92 Qwen3.6-PlusAlibaba
27.8%
$0.33 / $1.95
92 GPT-5-Codex (high reasoning)OpenAI
27.8%
–
92 Hy3 Preview (reasoning on)Tencent · best of 2 settings
27.8%
$0.18 / $0.60
92 GLM-5-TurboZ.ai
27.8%
$1.20 / $4
96 GLM-4.7Z.ai · best of 2 settings
27.4%
$0.60 / $2.20
97 MiMo-V2.5Xiaomi
27.2%
$0.40 / $2
98 Grok 4xAI
26.7%
–
99 GPT-5.1-Codex (high reasoning)OpenAI
25.7%
$1.25 / $10
100 Qwen3.5-122B-A10BAlibaba · best of 2 settings
25.2%
$0.26 / $2.08
101 DeepSeek-V3.2 (reasoning on)DeepSeek · best of 2 settings
24.6%
$0.28 / $0.42
102 Grok 4.20 (0309, non-reasoning, no reasoning)xAI
24.5%
–
103 Qwen3.5-27BAlibaba · best of 2 settings
23.9%
$0.27 / $2.16
104 Kimi K2 ThinkingMoonshot AI
23.8%
$0.60 / $2.50
105 Gemma 4 31BGoogle · best of 2 settings
23.6%
$0.15 / $0.40
106 MiniMax-M2.1MiniMax
23.2%
–
107 Qwen3.6-27BAlibaba · best of 2 settings
23.1%
$0.30 / $3.20
108 Gemini 2.5 ProGoogle
22.5%
$1.25 / $10
109 Qwen3.6-35B-A3BAlibaba · best of 2 settings
22.2%
$0.10 / $0.95
110 Muse Glimmer (high reasoning)Meta
22.0%
–
111 GPT-5.5 Instant (2026-05-26)OpenAI · best of 2 settings
21.6%
–
112 GPT-5 mini (high reasoning)OpenAI · best of 3 settings
21.5%
$0.25 / $2
113 Step 3.5 FlashStepFun
21.1%
$0.10 / $0.30
114 Qwen3.5-35B-A3BAlibaba · best of 2 settings
21.0%
$0.15 / $1
115 Nemotron 3 Super 120B A12BNVIDIA
20.8%
$0.08 / $0.45
116 MiniMax-M2.5MiniMax
20.5%
$0.29 / $1.20
117 o3OpenAI
20.1%
$2 / $8
118 gpt-oss-120b (high reasoning)OpenAI · best of 2 settings
19.6%
$0.10 / $0.50
119 Gemma 4 26B A4BGoogle · best of 2 settings
19.3%
$0.09 / $0.30
119 Grok 4.1 Fast (reasoning)xAI
19.3%
–
121 Grok 4 Fast (reasoning)xAI
19.1%
–
122 Gemini 3.5 Flash-LiteGoogle
18.8%
$0.30 / $2.50
123 GPT-5.1-Codex-Mini (high reasoning)OpenAI
18.5%
$0.25 / $2
124 Claude Sonnet 4.5 (reasoning on)Anthropic · best of 2 settings
17.8%
$3 / $15
125 Gemini 3.1 Flash-Lite PreviewGoogle
17.2%
$0.25 / $1.50
126 Mercury 2Inception
17.1%
–
126 GLM-5V-TurboZ.ai
17.1%
$1.20 / $4
128 o4-mini (high reasoning)OpenAI
16.5%
–
129 DeepSeek-V3.1-Terminus (reasoning on)DeepSeek · best of 2 settings
16.4%
$0.27 / $1
130 Qwen3-235B-A22B-Thinking-2507Alibaba
15.9%
$0.45 / $3.50
131 Trinity Large ThinkingArcee AI
15.8%
–
131 DeepSeek-R1-0528DeepSeek
15.8%
$0.50 / $2.18
133 Gemma 4 12BGoogle · best of 2 settings
15.7%
–
134 Qwen3.5-9BAlibaba · best of 2 settings
14.9%
$0.10 / $0.15
134 Qwen3.5-Omni-PlusAlibaba
14.9%
–
134 DeepSeek-V3.2-Exp (reasoning on)DeepSeek · best of 2 settings
14.9%
$0.27 / $0.41
137 GLM-4.6 (reasoning on)Z.ai · best of 2 settings
14.5%
$0.50 / $2
138 DeepSeek-V3.1 (reasoning on)DeepSeek · best of 2 settings
14.3%
$0.27 / $1
139 Gemini 2.5 Flash Preview (09-2025, reasoning on)Google · best of 2 settings
13.8%
–
139 Mistral Medium 3.5Mistral AI
13.8%
$1.50 / $7.50
141 MiniMax-M2MiniMax
13.7%
–
142 GLM-4.5Z.ai
13.0%
$0.60 / $2.20
143 Qwen3-Max-Thinking-PreviewAlibaba
12.7%
–
144 Qwen3-Next-80B-A3B-ThinkingAlibaba
12.6%
–
145 Claude Opus 4.1 (reasoning on)Anthropic
12.5%
$15 / $75
146 Gemini 2.5 Flash (reasoning on)Google · best of 2 settings
12.1%
$0.30 / $2.50
147 o3-mini (high reasoning)OpenAI · best of 2 settings
12.0%
–
148 Qwen3-MaxAlibaba
11.9%
–
148 Qwen3-VL-235B-A22B-ThinkingAlibaba
11.9%
–
150 Nova 2 Lite (high reasoning)Amazon · best of 4 settings
11.6%
$0.30 / $2.50
151 Nemotron 3 Nano 30B A3B (reasoning on)NVIDIA · best of 2 settings
11.4%
$0.06 / $0.24
152 Qwen3-235B-A22B-Instruct-2507Alibaba
11.1%
$0.15 / $0.75
153 Qwen3-235B-A22B (reasoning on)Alibaba · best of 2 settings
11.0%
–
153 gpt-oss-20b (high reasoning)OpenAI · best of 2 settings
11.0%
$0.03 / $0.14
155 DiffusionGemma 26B A4BGoogle
10.8%
–
156 Claude Sonnet 4 (reasoning on)Anthropic · best of 2 settings
10.7%
–
157 Claude Haiku 4.5 (reasoning on)Anthropic · best of 2 settings
10.4%
$1 / $5
158 Qwen3-30B-A3B-Thinking-2507Alibaba
10.3%
–
158 Magistral Medium 1.2Mistral AI
10.3%
–
158 Solar Pro 3Upstage
10.3%
–
161 Qwen3-Coder-NextAlibaba
10.1%
$0.12 / $0.80
161 Qwen3-Max-PreviewAlibaba
10.1%
–
161 Qwen3-VL-32B-ThinkingAlibaba
10.1%
–
164 Qwen3.5-4B (reasoning on)Alibaba · best of 2 settings
9.9%
–
164 Mistral Small 4Mistral AI · best of 2 settings
9.9%
$0.15 / $0.60
166 Granite 4.2 8BIBM
9.7%
$0.06 / $0.15
167 GLM-4.6V (reasoning on)Z.ai · best of 2 settings
9.6%
$0.30 / $0.90
168 GPT-5 nano (high reasoning)OpenAI · best of 3 settings
9.5%
$0.05 / $0.40
169 Qwen3-VL-30B-A3B-ThinkingAlibaba
8.9%
$0.29 / $1
170 DeepSeek-R1DeepSeek
8.5%
–
171 Grok Code Fast 1xAI
8.0%
–
172 Qwen3.5-Omni-FlashAlibaba
7.6%
–
172 Qwen3-Next-80B-A3B-InstructAlibaba
7.6%
$0.09 / $1.10
172 GLM-4.7-FlashZ.ai · best of 2 settings
7.6%
$0.06 / $0.40
175 Qwen3-Omni-30B-A3B-ThinkingAlibaba
7.5%
–
176 Qwen3-32B (reasoning on)Alibaba · best of 2 settings
7.4%
$0.14 / $0.40
176 Kimi K2 (0711)Moonshot AI
7.4%
$0.57 / $2.30
178 Gemini 2.5 Flash-Lite Preview (09-2025, reasoning on)Google · best of 2 settings
7.0%
–
178 o1OpenAI
7.0%
–
178 GLM-4.5-AirZ.ai
7.0%
$0.14 / $0.86
181 Qwen3-30B-A3B-Instruct-2507Alibaba
6.9%
$0.09 / $0.30
182 Qwen3-VL-32B-InstructAlibaba
6.8%
–
182 Gemini 2.5 Flash-Lite (reasoning on)Google · best of 2 settings
6.8%
$0.10 / $0.40
184 Qwen3-VL-235B-A22B-InstructAlibaba
6.6%
$0.21 / $1.90
184 GPT-5 ChatOpenAI
6.6%
–
186 Magistral Small 1.2Mistral AI
6.4%
–
186 Kimi K2 (0905)Moonshot AI
6.4%
$0.60 / $2.50
188 Qwen3-VL-30B-A3B-InstructAlibaba
6.3%
$0.20 / $0.70
188 GLM-4.5V (reasoning on)Z.ai · best of 2 settings
6.3%
$0.60 / $1.80
190 Qwen3-30B-A3B (reasoning on)Alibaba · best of 2 settings
6.2%
$0.12 / $0.50
190 Qwen3-4B-Thinking-2507Alibaba
6.2%
–
192 Llama 3.2 1B InstructMeta
5.5%
$0.027 / $0.20
193 Ministral 3 3BMistral AI
5.4%
$0.10 / $0.10
194 Gemma 3 4BGoogle
5.3%
$0.04 / $0.05
194 Llama 3.1 8B InstructMeta
5.3%
$0.05 / $0.08
194 Llama 3.2 3B InstructMeta
5.3%
$0.05 / $0.33
197 Qwen3.5-0.8B (no reasoning)Alibaba · best of 2 settings
5.1%
–
197 Grok 4.1 Fast (non-reasoning, no reasoning)xAI
5.1%
–
199 Qwen3.5-2B (no reasoning)Alibaba · best of 2 settings
5.0%
–
199 GPT-4.1 miniOpenAI
5.0%
$0.40 / $1.60
201 Llama 4 MaverickMeta
4.9%
$0.35 / $1
202 Gemma 4 E2BGoogle · best of 2 settings
4.8%
–
202 Gemma 4 E4B (no reasoning)Google · best of 2 settings
4.8%
–
204 DeepSeek-V3-0324DeepSeek
4.7%
$0.25 / $1
204 Mistral Medium 3.1Mistral AI
4.7%
$0.40 / $2
206 Qwen3-Omni-30B-A3B-InstructAlibaba
4.6%
–
206 Qwen3-VL-4B-ThinkingAlibaba
4.6%
–
206 Nova Micro 1.0Amazon
4.6%
$0.035 / $0.14
206 Ministral 3 14BMistral AI
4.6%
$0.20 / $0.20
210 Qwen3-14B (reasoning on)Alibaba · best of 2 settings
4.5%
$0.10 / $0.22
210 Qwen3-4B-Instruct-2507Alibaba
4.5%
–
210 Qwen3-Coder-480B-A35BAlibaba
4.5%
$0.30 / $1
210 Llama 3.1 70B InstructMeta
4.5%
$0.72 / $0.72
210 Grok 4 Fast (non-reasoning, no reasoning)xAI
4.5%
–
215 Gemma 3 27BGoogle
4.4%
$0.08 / $0.35
215 Reka Flash 3Reka AI
4.4%
$0.10 / $0.20
217 Nova Lite 1.0Amazon
4.3%
$0.06 / $0.24
217 Ministral 3 8BMistral AI
4.3%
$0.15 / $0.15
217 Mistral Small 3.1 24BMistral AI
4.3%
$0.35 / $0.56
217 Mistral Small 3.2 24BMistral AI
4.3%
$0.094 / $0.25
221 Gemma 3 12BGoogle
4.2%
$0.05 / $0.15
221 Mistral Large 3Mistral AI
4.2%
$0.50 / $1.50
221 GPT-4.1OpenAI
4.2%
$2 / $8
221 GPT-4o mini (2024-07-18)OpenAI
4.2%
$0.15 / $0.60
225 Mistral Medium 3Mistral AI
4.1%
$0.40 / $2
226 Command ACohere
4.0%
$2.50 / $10
226 Mixtral 8x22B InstructMistral AI
4.0%
$2 / $6
228 Qwen3-8B (reasoning on)Alibaba · best of 2 settings
3.9%
–
229 Qwen3-Coder-30B-A3B-InstructAlibaba
3.8%
$0.07 / $0.28
229 Qwen3-VL-8B-ThinkingAlibaba
3.8%
–
229 Llama 4 ScoutMeta
3.8%
$0.10 / $0.30
229 Phi-4Microsoft
3.8%
$0.07 / $0.14
229 Devstral MediumMistral AI
3.8%
–
229 Devstral Small 1.1Mistral AI
3.8%
–
229 Mistral Small 3Mistral AI
3.8%
$0.05 / $0.08
229 GPT-4.1 nanoOpenAI
3.8%
$0.10 / $0.40
237 Qwen2.5-72B-InstructAlibaba
3.6%
$0.36 / $0.40
237 Qwen3-VL-4B-InstructAlibaba
3.6%
–
237 Gemma 3 270MGoogle
3.6%
–
237 Llama 3.3 70B InstructMeta
3.6%
$0.71 / $0.71
237 Devstral 2Mistral AI
3.6%
$0.40 / $2
242 Qwen2.5-Coder-32B-InstructAlibaba
3.5%
$0.66 / $1
242 Devstral Small 2Mistral AI
3.5%
–
244 Nova Pro 1.0Amazon
3.2%
$0.80 / $3.20
245 DeepSeek-V3DeepSeek
2.9%
$0.26 / $1.03
245 Mistral Large 2 (2407)Mistral AI
2.9%
$2 / $6
247 Qwen3-VL-8B-InstructAlibaba
2.7%
$0.25 / $0.75
248 Kimi Linear 48B A3B InstructMoonshot AI
2.5%
–
249 GPT-4o (2024-08-06)OpenAI
2.3%
$2.50 / $10
250 GPT-4o (2024-05-13)OpenAI
1.8%
$5 / $15

Swipe the table sideways for more columns.

Source: Artificial Analysis, via Spring Prompt, springprompt.com/benchmarks/artificial-analysis. Test your own prompt on these models →

Ranks follow the score as shown, so equal numbers share a rank. Each model is shown at its best setting; show every setting. Results as published by Artificial Analysis; we do not re-run them.

Progress over time

Humanity's Last Exam, plotted at each model's release date, higher is better. Each model counts once, at its best setting. Scores are today's, not launch day's. Hover or tap a mark for its model.

Best score so far

Every model at its release date. The line moves each time a new release beats every earlier one, and each new best is named where there is room.

-20%0%20%40%60%80% Jul 20242025Jul2026Jul GPT-4o mini (2024-07-18)Llama 3.1 8B Instructo1DeepSeek-R1o3-miniGemini 2.5 ProGrok 4GPT-5Gemini 3 Pro PreviewGPT-5.3-CodexGemini 3.1 Pro PreviewClaude Opus 4.8Claude Fable 5Claude Fable 5.1Claude Opus 5.5

Best score by lab

Each lab's best so far, stepping whenever one of its releases beats its own earlier models. The 6 labs with the best scores today; 19 others not drawn.

AnthropicGoogleOpenAIMoonshot AIMetaXiaomi

-20%0%20%40%60%80% Jul 20242025Jul2026Jul AnthropicGoogleOpenAIMoonshot AIMetaXiaomi

What it measures

Share of expert-written exam questions answered correctly, as run by Artificial Analysis with its own harness and prompts.

What it does not measure

Not everyday work; academic questions at the edge of expertise.

291 results from Artificial Analysis not ranked here · show why

We rank a result only when we can tie it to a specific model you can use. These are left out:

  • Not on sale through the API providers we track: 276
  • A different snapshot or variant from the model we list: 14
  • An unusual combination of settings: 1

Data sourced from Artificial Analysis. Licence: Artificial Analysis commercial data licence.