Skip to content
IFAI

Model rankings

285 models · 2 boards · 2 sources

Artificial AnalysisIntelligence IndexArenaText arena

Text arena · Anonymous head-to-head votes · Published March 20, 2025

Most preferred
Best open model
deepseek-r1DeepSeek1,317Score
Best from China
deepseek-r1DeepSeek1,317Score
Most voted
claude-3-opus-20240229Anthropic202,722Votes
Top three by Score
3o1-2024-12-17OpenAI1,321
Top three open models
1deepseek-r1DeepSeek · #4 overall1,317
2gemma-3-27b-itGoogle · #12 overall1,279
3deepseek-v3DeepSeek · #13 overall1,277
This week's changes
QwQ 32BIntelligence · #4New
Reka Flash 3Intelligence · #12New
Command AIntelligence · #26New
Gemma 3 27B InstructIntelligence · #30New
RankChangeModelScore95% intervalVotesPrice $/M
1—gpt-4.5-preview-2025-02-27ClosedOpenAI1,357±69,871—
2—grok-3-preview-02-24ClosedxAI1,329±69,871—
3—o1-2024-12-17ClosedOpenAI1,321±424,350—
4Newclaude-3-7-sonnet-20250219-thinking-32kClosedAnthropic1,315±94,487—
5—o1-previewClosedOpenAI1,303±333,186—
6▼ 2claude-3-7-sonnet-20250219ClosedAnthropic1,302±79,503—
7▼ 1o3-mini-highClosedOpenAI1,286±613,615—
8—qwen2.5-maxClosedAlibaba1,286±516,493—
9▼ 2claude-3-5-sonnet-20241022ClosedAnthropic · 1 more variant1,286±362,264—
10▼ 1gemini-2.0-flash-001ClosedGoogle1,284±417,689—
11▼ 1gemini-1.5-pro-002ClosedGoogle1,269±358,685—
12▼ 1o3-miniClosedOpenAI1,267±520,120—
13▼ 1gemini-2.0-flash-lite-preview-02-05ClosedGoogle1,267±417,387—
14—gpt-4o-2024-05-13ClosedOpenAI1,263±2117,780—
15▼ 2hunyuan-turbos-20250226ClosedTencent1,263±122,330—
16▼ 1qwen-plus-0125ClosedAlibaba1,262±76,061—
17▼ 1hunyuan-turbo-0110ClosedTencent1,261±102,390—
18▼ 1glm-4-plus-0111ClosedZhipu · 1 more variant1,256±76,039—
19▼ 1o1-miniClosedOpenAI1,255±354,982—
20▼ 1gemini-advanced-0514ClosedGoogle1,254±352,159—
21▼ 1step-2-16k-exp-202412ClosedStepFun1,251±75,130—
22▼ 1gpt-4o-2024-08-06ClosedOpenAI1,250±347,990—
23▼ 1grok-2-2024-08-13ClosedxAI1,249±367,111—
24▼ 1yi-lightningClosed01 AI1,246±328,970—
25▼ 1gpt-4-turbo-2024-04-09ClosedOpenAI1,242±2102,191—
26▼ 1hunyuan-large-2025-02-10ClosedTencent1,242±83,859—
27▼ 1gemini-1.5-pro-001ClosedGoogle1,241±382,467—
28▼ 1claude-3-opus-20240229ClosedAnthropic1,239±2202,722—
29▼ 1gpt-4-1106-previewClosedOpenAI1,235±2103,769—
30▼ 1gpt-4o-mini-2024-07-18ClosedOpenAI1,234±369,078—
31▼ 1claude-3-5-haiku-20241022ClosedAnthropic1,232±428,471—
32▼ 1qwen2.5-plus-1127ClosedAlibaba1,231±510,723—
33▼ 1gpt-4-0125-previewClosedOpenAI1,231±297,097—
34▼ 1gemini-1.5-flash-002ClosedGoogle1,230±337,020—
35▼ 1hunyuan-standard-2025-02-10ClosedTencent1,228±104,019—
36▼ 1grok-2-mini-2024-08-13ClosedxAI1,224±355,462—
37▼ 1amazon-nova-pro-v1.0ClosedAmazon1,209±421,862—
38▼ 1reka-core-20240904ClosedReka AI1,207±67,940—
39▼ 1gemini-1.5-flash-001ClosedGoogle1,204±365,668—
40▼ 1gpt-4-0314ClosedOpenAI1,200±355,982—
41▼ 1claude-3-sonnet-20240229ClosedAnthropic1,197±2113,075—
42▼ 1reka-flash-20240904ClosedReka AI1,193±68,136—
43▼ 1glm-4-0520ClosedZhipu AI1,188±610,215—
44▼ 1gpt-4-0613ClosedOpenAI1,185±391,631—
45▼ 1amazon-nova-lite-v1.0ClosedAmazon1,181±419,865—
46▼ 1gemini-1.5-flash-8b-001ClosedGoogle1,180±337,691—
47▼ 1claude-3-haiku-20240307ClosedAnthropic1,178±2122,347—
48▼ 1amazon-nova-micro-v1.0ClosedAmazon1,163±519,908—
49▼ 1mistral-large-2402ClosedMistral1,158±364,933—
50▼ 1hunyuan-standard-256kClosedTencent1,153±92,904—

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 “—”.