Skip to content
IFAI

Model rankings

391 models · 2 boards · 2 sources

Artificial AnalysisIntelligence IndexArenaText arena

Intelligence Index · A composite of several evaluations · Updated May 27, 2025

Smartest
o4-mini (high)OpenAI69.8Intelligence
Best open model
Qwen3 235B A22B (Reasoning)Alibaba62.3Intelligence
Best value
Gemma 3 4B InstructGoogle979 pts/$Value
Best from China
Qwen3 235B A22B (Reasoning)Alibaba62.3Intelligence
Top three by Intelligence
1o4-mini (high)OpenAI69.8
3o3OpenAI66.8
Top three open models
1Qwen3 235B A22B (Reasoning)Alibaba · #7 overall62.3
3DeepSeek R1DeepSeek · #11 overall60.2
This week's changes
Claude 4 OpusIntelligence · #14New
Grok 3 Reasoning BetaIntelligence · #16New
o1-proIntelligence · #16New
GPT-4.5 (Preview)Intelligence · #22New
RankChangeModelIntelligencePrice $/MSpeed tok/sFirst token
1—o4-mini (high)ClosedOpenAI69.8$1.9——
2—Gemini 2.5 Pro Preview (Mar' 25)ClosedGoogle68.9$3.4——
3—o3ClosedOpenAI66.8$18——
4—Grok 3 mini Reasoning (high)ClosedxAI66.7$0.35——
5—o3-mini (high)ClosedOpenAI65.5$1.9——
6NewGemini 2.5 Flash Preview (May '25) (Reasoning)ClosedGoogle65.1$0.99——
7▼ 1o3-miniClosedOpenAI62.9$1.9——
8▼ 1Qwen3 235B A22B (Reasoning)OpenAlibaba62.3$2.6——
9▼ 1o1ClosedOpenAI61.9$26——
10▼ 1Llama 3.1 Nemotron Ultra 253B v1 (Reasoning)OpenNVIDIA60.8$0.90——
11NewClaude 4 Sonnet (Extended Thinking)ClosedAnthropic60.8$6.0——
12▼ 2Gemini 2.5 Flash Preview (Reasoning)ClosedGoogle60.5$0.99——
13▼ 2DeepSeek R1OpenDeepSeek60.2$0.96——
14Newo1-previewClosedOpenAI59.6$26——
15▼ 3Qwen3 32B (Reasoning)OpenAlibaba59.2$2.6——
16▼ 3QwQ 32BOpenAlibaba58.1$0.43——
17NewClaude 4 OpusClosedAnthropic57.7$30——
18▼ 4Claude 3.7 Sonnet (Extended Thinking)ClosedAnthropic57.4$6.0——
19NewGrok 3 Reasoning BetaClosedxAI56.1$0.000——
19Newo1-proClosedOpenAI56.1$263——
21▼ 6Qwen3 14B (Reasoning)OpenAlibaba55.7$1.3——
22▼ 6Qwen3 30B A3B (Reasoning)OpenAlibaba55.6$0.75——
23▼ 6o1-miniClosedOpenAI53.8$1.9——
24NewGemini 2.5 Flash Preview (May '25)ClosedGoogle53.4$0.26——
25▼ 7DeepSeek V3 0324 (Mar' 25)OpenDeepSeek53.2$0.48——
26NewClaude 4 SonnetClosedAnthropic53.0$6.0——
26NewGPT-4.5 (Preview)ClosedOpenAI53.0$94——
28▼ 9GPT-4.1 miniClosedOpenAI52.9$0.70——
29▼ 9GPT-4.1ClosedOpenAI52.6$3.5——
30▼ 9Gemini 2.0 Flash Thinking Experimental (Jan '25)ClosedGoogle52.3$0.000——
31▼ 9DeepSeek R1 Distill Qwen 32BOpenDeepSeek51.9$0.22——
32▼ 9Qwen3 8B (Reasoning)OpenAlibaba51.4$0.66——
33▼ 9Llama 3.3 Nemotron Super 49B v1 (Reasoning)OpenNVIDIA51.2$0.000——
34▼ 9Grok 3ClosedxAI50.6$6.0——
35▼ 9Llama 4 MaverickOpenMeta50.5$0.37——
36▼ 9GPT-4o (March 2025, chatgpt-4o-latest)ClosedOpenAI50.3$7.5——
37▼ 9Gemini 2.0 Pro Experimental (Feb '25)ClosedGoogle49.2$0.000——
38▼ 9DeepSeek R1 Distill Qwen 14BOpenDeepSeek49.0$0.88——
39▼ 9Mistral Medium 3ClosedMistral49.0$0.80——
40NewSonar ReasoningClosedPerplexity48.8$2.0——
41▼ 10Gemini 2.5 Flash PreviewClosedGoogle48.8$0.26——
42▼ 10DeepSeek R1 Distill Llama 70BOpenDeepSeek48.2$0.77——
43▼ 10Claude 3.7 Sonnet (Standard)ClosedAnthropic48.2$6.0——
44▼ 10Gemini 2.0 Flash (Feb '25)ClosedGoogle48.1$0.18——
45▼ 10Qwen3 4B (Reasoning)OpenAlibaba47.3$0.40——
46▼ 10Reka Flash 3OpenReka AI47.1$0.35——
47▼ 10Qwen3 235B A22BOpenAlibaba47.1$1.2——
48▼ 10Gemini 2.0 Flash (experimental)ClosedGoogle46.4$0.000——
49▼ 10DeepSeek V3 (Dec '24)OpenDeepSeek45.6$0.48——
50▼ 10Qwen2.5 MaxClosedAlibaba45.3$2.8——

Intelligence over time

Each maker's best model at the time on the Artificial Analysis Intelligence Index, by the month it was released.

Gained most · Anthropic +57.2Highest now · Anthropic 64.6Gap between highest and lowest 18.1 ptsTime range Nov 2025 – Nov 2026
64.6Anthropic60.0Google56.9OpenAI50.1Meta46.5Kimi

Arena data comes from the leaderboard dataset it publishes on Hugging Face (CC BY 4.0). Only each board's overall leaderboard is used; scores and ranks are unchanged. dataset · CC BY 4.0

Artificial Analysis data comes from its API (artificialanalysis.ai). Scores are unchanged; ranks follow the scores. artificialanalysis.ai

Open or closed follows the license Arena gives; prices and Artificial Analysis intelligence show only where model names match, otherwise “—”.