Skip to content
IFAI

Model rankings

394 models · 2 boards · 2 sources

Artificial AnalysisIntelligence IndexArenaText arena

Intelligence Index · A composite of several evaluations · Updated June 3, 2025

Smartest
o3OpenAI70.0Intelligence
Best open model
DeepSeek R1 0528 (May '25)DeepSeek68.3Intelligence
Best value
Gemma 3 4B InstructGoogle979 pts/$Value
Best from China
DeepSeek R1 0528 (May '25)DeepSeek68.3Intelligence
Top three by Intelligence
1o3OpenAI70.0
2o4-mini (high)OpenAI69.8
Top three open models
1DeepSeek R1 0528 (May '25)DeepSeek · #4 overall68.3
2Qwen3 235B A22B (Reasoning)Alibaba · #9 overall62.3
This week's changes
DeepSeek R1 0528 (May '25)Intelligence · #4Up 7
Claude 4 Opus (Extended Thinking)Intelligence · #8Up 6
o3Intelligence · #1Up 2
RankChangeModelIntelligencePrice $/MSpeed tok/sFirst token
1▲ 2o3ClosedOpenAI70.0$18——
2▼ 1o4-mini (high)ClosedOpenAI69.8$1.9——
3▼ 1Gemini 2.5 Pro Preview (Mar' 25)ClosedGoogle68.9$3.4——
4NewDeepSeek R1 0528 (May '25)OpenDeepSeek68.3$0.96——
5NewGemini 2.5 Pro Preview (May' 25)ClosedGoogle67.9$3.4——
6▼ 2Grok 3 mini Reasoning (high)ClosedxAI66.7$0.35——
7▼ 2o3-mini (high)ClosedOpenAI65.5$1.9——
8▼ 2Gemini 2.5 Flash Preview (May '25) (Reasoning)ClosedGoogle65.1$0.99——
9NewClaude 4 Opus (Extended Thinking)ClosedAnthropic64.4$30——
10▼ 3o3-miniClosedOpenAI62.9$1.9——
11▼ 3Qwen3 235B A22B (Reasoning)OpenAlibaba62.3$2.6——
12▼ 3o1ClosedOpenAI61.9$26——
13▼ 3Llama 3.1 Nemotron Ultra 253B v1 (Reasoning)OpenNVIDIA60.8$0.90——
14▼ 3Claude 4 Sonnet (Extended Thinking)ClosedAnthropic60.8$6.0——
15▼ 3Gemini 2.5 Flash Preview (Reasoning)ClosedGoogle60.5$0.99——
16NewDeepSeek R1 (Jan '25)OpenDeepSeek60.2$2.6——
17▼ 3o1-previewClosedOpenAI59.6$26——
18▼ 3Qwen3 32B (Reasoning)OpenAlibaba59.2$2.6——
19▼ 3QwQ 32BOpenAlibaba58.1$0.54——
20▼ 3Claude 4 OpusClosedAnthropic57.7$30——
21▼ 3Claude 3.7 Sonnet (Extended Thinking)ClosedAnthropic57.4$6.0——
22▼ 3Grok 3 Reasoning BetaClosedxAI56.1$0.000——
22▼ 3o1-proClosedOpenAI56.1$263——
24▼ 3Qwen3 14B (Reasoning)OpenAlibaba55.7$1.3——
25▼ 3Qwen3 30B A3B (Reasoning)OpenAlibaba55.6$0.75——
26▼ 3o1-miniClosedOpenAI53.8$1.9——
27▼ 3Gemini 2.5 Flash Preview (May '25)ClosedGoogle53.4$0.26——
28NewDeepSeek V3 0324 (Mar '25)OpenDeepSeek53.2$0.48——
29▼ 3Claude 4 SonnetClosedAnthropic53.0$6.0——
29▼ 3GPT-4.5 (Preview)ClosedOpenAI53.0$94——
31▼ 3GPT-4.1 miniClosedOpenAI52.9$0.70——
32▼ 3GPT-4.1ClosedOpenAI52.6$3.5——
33▼ 3Gemini 2.0 Flash Thinking Experimental (Jan '25)ClosedGoogle52.3$0.000——
34▼ 3DeepSeek R1 Distill Qwen 32BOpenDeepSeek51.9$0.22——
35▼ 3Qwen3 8B (Reasoning)OpenAlibaba51.4$0.66——
36▼ 3Llama 3.3 Nemotron Super 49B v1 (Reasoning)OpenNVIDIA51.2$0.000——
37▼ 3Grok 3ClosedxAI50.6$6.0——
38▼ 3Llama 4 MaverickOpenMeta50.5$0.37——
39▼ 3GPT-4o (March 2025, chatgpt-4o-latest)ClosedOpenAI50.3$7.5——
40▼ 3Gemini 2.0 Pro Experimental (Feb '25)ClosedGoogle49.2$0.000——
41▼ 3DeepSeek R1 Distill Qwen 14BOpenDeepSeek49.0$0.88——
42▼ 3Mistral Medium 3ClosedMistral49.0$0.80——
43▼ 3Sonar ReasoningClosedPerplexity48.8$2.0——
44▼ 3Gemini 2.5 Flash PreviewClosedGoogle48.8$0.26——
45▼ 3DeepSeek R1 Distill Llama 70BOpenDeepSeek48.2$0.77——
46▼ 3Claude 3.7 Sonnet (Standard)ClosedAnthropic48.2$6.0——
47▼ 3Gemini 2.0 Flash (Feb '25)ClosedGoogle48.1$0.18——
48▼ 3Qwen3 4B (Reasoning)OpenAlibaba47.3$0.40——
49▼ 3Reka Flash 3OpenReka AI47.1$0.35——
50▼ 3Qwen3 235B A22BOpenAlibaba47.1$1.2——

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 “—”.