Benchmark · KnowledgeCompetitive

Artificial Analysis · GDPval

Updated 2026-10-15
Models tested
46
Top score
68.3
Claude Opus 5.5
Median
41.5
min 0.7
Top-5 spread
σ 3.5
Competitive

Best score over time · one chart, every benchmark

ARTIFICIAL ANALYSIS · GDPVAL46 MODELS · FRONTIER RUNNING MAX0255075100SCORE ↑Aug 25Nov 25Mar 26Jun 26Oct 26RELEASE DATE →benchgecko.ai/benchmark/aa-gdpval · frontier
Frontier on Artificial Analysis · GDPval rose from 5.4 to 68.3 in 14 months · +62.9 points · latest leader Claude Opus 5.5 from Anthropic.
Pink dots = frontier records · 9 totalClick to open model page

46 models tested · sorted by score

#ModelScore
1Anthropic logoClaude Opus 5.568.3
2Anthropic logoClaude Sonnet 5.566.9
3Anthropic logoClaude Fable 5.162.7
4xAI logoGrok 4.760.6
5xiaomi logoMiMo-V2.6-Pro59.2
6Alibaba Qwen logoQwen3.8 Max (0902)58.3
7z-ai logoGLM 5.357.5
8z-ai logoGLM 5.3 Flash57.2
9Google DeepMind logoGemini 4 Argon56.2
10Anthropic logoClaude Haiku 5.555.9
11xiaomi logoMiMo-V2.6-Flash55.4
12Alibaba Qwen logoQwen3.8 2.4T A95B55.4
13DeepSeek logoDeepSeek V4.1 Flash55.0
14OpenAI logoGPT-6.1 Sol54.7
15stepfun logoStep 5 Preview54.3
16OpenAI logoGPT-6 Astra53.6
17moonshotai logoKimi K351.7
18OpenAI logoGPT-5.6 Terra47.6
19DeepSeek logoDeepSeek V4 Pro47.5
20Google DeepMind logoGemini 3.8 Flash46.8
21OpenAI logoGPT-6 Luna46.6
22Mistral AI logoMistral Large 446.1
23Alibaba Qwen logoQwen3.8 27B46.0
24minimax logoMiniMax M337.1
25NVIDIA logoNemotron 3.5 Super VL33.1
26upstage logoSolar Pro 430.3
27upstage logoSolar Mini 429.5
28tencent logoHy328.3
29DeepSeek logoDeepSeek V4 Pro 081327.9
30moonshotai logoKimi K2.7 Code26.9
31NVIDIA logoNemotron 3 Ultra25.7
32Google DeepMind logoGemini 3.5 Flash Lite24.2
33Alibaba Qwen logoQwen3.6 35B A3B19.6
34meituan logoLongCat 2.018.5
35Google DeepMind logoGemini 3.1 Pro Preview14.6
36Alibaba Qwen logoQwen3.5 397B A17B14.6
37Alibaba Qwen logoQwen3.7 Plus13.4
38Mistral AI logoMistral Medium 3.513.0
39Alibaba Qwen logoQwen3.5-122B-A10B10.5
40NVIDIA logoNemotron 3.5 Lightning6.9
41Google DeepMind logoGemma 4 31B6.0
42OpenAI logogpt-oss-120b5.4
43ibm-granite logoGranite 4.2 8B3.8
44Google DeepMind logoGemma 4 26B A4B 3.3
45Cohere logoCommand A+0.7
46Alibaba Qwen logoQwen3 Coder Next0.7

What it measures · how it is scored · sources · as of 2026-11-17

  • How it is scored · Scored as a percentage · higher is better · maximum 100
  • Scores from · Artificial Analysis (46 scores)

Same category · related evaluations