Skip to content
IFAI

Model rankings

539 models · 3 boards · 2 sources

Artificial AnalysisIntelligence IndexArenaText arena

Intelligence Index · A composite of several evaluations · Updated September 23, 2025

Smartest
GPT-5 (high)OpenAI66.7Intelligence
Best open model
gpt-oss-120B (high)OpenAI57.9Intelligence
Best value
Gemma 3n E4B InstructGoogle624 pts/$Value
Best from China
Qwen3 235B A22B 2507 (Reasoning)Alibaba57.5Intelligence
Top three by Intelligence
1GPT-5 (high)OpenAI66.7
2Grok 4xAI65.3
3o3-proOpenAI65.3
Top three open models
1gpt-oss-120B (high)OpenAI · #10 overall57.9
2Qwen3 235B A22B 2507 (Reasoning)Alibaba · #11 overall57.5
3Qwen3 Next 80B A3B (Reasoning)Alibaba · #14 overall54.3
This week's changes
Grok 4 Fast (Reasoning)Intelligence · #6New
Magistral Medium 1.2Intelligence · #18New
Magistral Small 1.2Intelligence · #39New
Magistral Medium 1Intelligence · #69New
Magistral Small 1Intelligence · #73New
RankChangeModelIntelligencePrice $/MSpeed tok/sFirst token
1—gpt-oss-120B (high)OpenOpenAI57.9$0.26——
2—Qwen3 235B A22B 2507 (Reasoning)OpenAlibaba · 3 more variants57.5$2.6——
3—Qwen3 Next 80B A3B (Reasoning)OpenAlibaba · 1 more variant54.3$1.9——
4—DeepSeek V3.1 (Reasoning)OpenDeepSeek · 1 more variant54.0$0.45——
5—DeepSeek R1 0528 (May '25)OpenDeepSeek · 1 more variant52.0$1.4——
6—Kimi K2 0905OpenMoonshot AI · 1 more variant50.4$1.4——
7—GLM-4.5OpenZ AI49.4$0.99——
8—GLM-4.5-AirOpenZ AI48.8$0.43——
9—Qwen3 30B A3B 2507 (Reasoning)OpenAlibaba · 3 more variants46.4$0.75——
10—MiniMax M1 80kOpenMiniMax46.2$0.83——
11—Llama Nemotron Super 49B v1.5 (Reasoning)OpenNVIDIA · 1 more variant45.2$0.18——
12—gpt-oss-20B (high)OpenOpenAI44.8$0.088——
13—Qwen3 4B 2507 (Reasoning)OpenAlibaba · 2 more variants43.4$0.000——
14NewMagistral Small 1.2OpenMistral43.0$0.75——
15▼ 1EXAONE 4.0 32B (Reasoning)OpenLG AI Research · 1 more variant42.6$0.70——
16▼ 1Qwen3 Coder 480B A35BOpenAlibaba42.3$3.0——
17▼ 1MiniMax M1 40kOpenMiniMax41.6$0.83——
18▼ 1Hermes 4 - Llama-3.1 405B (Reasoning)OpenNous Research · 1 more variant41.6$1.5——
19▼ 1DeepSeek V3 0324OpenDeepSeek · 1 more variant41.3$0.99——
20▼ 1Hermes 4 - Llama-3.1 70B (Reasoning)OpenNous Research · 1 more variant39.2$0.20——
21▼ 1Qwen3 32B (Reasoning)OpenAlibaba · 1 more variant38.7$2.6——
22▼ 1Llama 3.1 Nemotron Ultra 253B v1 (Reasoning)OpenNVIDIA38.5$0.90——
23▼ 1NVIDIA Nemotron Nano 9B V2OpenNVIDIA · 1 more variant38.1$0.070——
24▼ 1QwQ 32BOpenAlibaba37.9$0.44——
25▼ 1GLM-4.5V (Reasoning)OpenZ AI · 1 more variant37.0$0.85——
26▼ 1Qwen3 14B (Reasoning)OpenAlibaba · 1 more variant36.0$1.3——
27▼ 1Llama 4 MaverickOpenMeta35.8$0.39——
28▼ 1Llama 3.3 Nemotron Super 49B v1 (Reasoning)OpenNVIDIA · 1 more variant35.5$0.000——
29▼ 1DeepSeek R1 0528 Qwen3 8BOpenDeepSeek34.8$0.065——
30▼ 1Qwen3 Coder 30B A3BOpenAlibaba33.4$0.90——
31▼ 1DeepSeek R1 Distill Qwen 32BOpenDeepSeek32.7$0.29——
32▼ 1Reka Flash 3OpenReka AI32.5$0.35——
33NewMagistral Small 1OpenMistral31.9$0.75——
34▼ 1DeepSeek R1 Distill Llama 70BOpenDeepSeek30.8$0.77——
35▼ 1DeepSeek R1 Distill Qwen 14BOpenDeepSeek29.7$0.88——
36▼ 1Mistral Small 3.2OpenMistral29.1$0.15——
37▲ 4Qwen2.5 Instruct 72BOpenAlibaba29.0$0.000——
38▼ 2Qwen3 8B (Reasoning)OpenAlibaba · 1 more variant28.3$0.66——
39▼ 2Llama 4 ScoutOpenMeta28.1$0.26——
40▼ 2Command AOpenCohere28.1$4.4——
41▼ 2QwQ 32B-PreviewOpenAlibaba28.0$0.67——
42▼ 2Llama 3.3 Instruct 70BOpenMeta27.9$0.57——
43▲ 19Devstral Small (Jul '25)OpenMistral · 1 more variant27.2$0.15——
44▲ 2Mistral Large 2 (Nov '24)OpenMistral · 1 more variant26.8$3.0——
45▼ 3Exaone 4.0 1.2B (Reasoning)OpenLG AI Research · 1 more variant26.7$0.000——
46▼ 3Llama 3.1 Nemotron Nano 4B v1.1 (Reasoning)OpenNVIDIA26.1$0.000——
47▼ 3MiniMax-Text-01OpenMiniMax25.7$0.43——
48▼ 3Llama 3.1 Instruct 405BOpenMeta25.7$3.5——
49▼ 2Llama 3.1 Tulu3 405BOpenAllen Institute for AI25.4$0.000——
50▼ 2Pixtral LargeOpenMistral25.0$3.0——

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 “—”.