Skip to content
IFAI

Model rankings

277 models · 2 boards · 2 sources

Artificial AnalysisIntelligence IndexArenaText arena

Text arena · Anonymous head-to-head votes · Published March 15, 2025

Most preferred
Best open model
deepseek-r1DeepSeek1,317Score
Best from China
deepseek-r1DeepSeek1,317Score
Most voted
claude-3-opus-20240229Anthropic202,714Votes
Top three by Score
3o1-2024-12-17OpenAI1,322
Top three open models
1deepseek-r1DeepSeek · #4 overall1,317
2deepseek-v3DeepSeek · #11 overall1,276
3gemma-3-27b-itGoogle · #12 overall1,276
This week's changes
SonarIntelligence · #15New
Sonar ProIntelligence · #16New
Nova LiteIntelligence · #42New
Jamba 1.6 LargeIntelligence · #45New
Phi-4 Multimodal InstructIntelligence · #62New
RankChangeModelScore95% intervalVotesPrice $/M
1—gpt-4.5-preview-2025-02-27ClosedOpenAI1,358±68,297—
2—grok-3-preview-02-24ClosedxAI1,330±78,926—
3—o1-2024-12-17ClosedOpenAI1,322±423,258—
4—claude-3-7-sonnet-20250219ClosedAnthropic1,304±68,411—
5—o1-previewClosedOpenAI1,303±433,185—
6—o3-mini-highClosedOpenAI1,287±612,604—
7▲ 1claude-3-5-sonnet-20241022ClosedAnthropic · 1 more variant1,285±461,714—
8▲ 1qwen2.5-maxClosedAlibaba1,285±515,411—
9▼ 2gemini-2.0-flash-001ClosedGoogle1,285±416,645—
10—gemini-1.5-pro-002ClosedGoogle1,268±358,691—
11—o3-miniClosedOpenAI1,268±519,024—
12—gemini-2.0-flash-lite-preview-02-05ClosedGoogle1,266±516,354—
13Newhunyuan-turbos-20250226ClosedTencent1,265±141,443—
14▼ 1gpt-4o-2024-05-13ClosedOpenAI1,263±2117,781—
15▼ 1qwen-plus-0125ClosedAlibaba1,262±76,058—
16Newhunyuan-turbo-0110ClosedTencent1,259±121,473—
17▼ 2glm-4-plus-0111ClosedZhipu · 1 more variant1,256±86,041—
18▼ 2o1-miniClosedOpenAI1,255±354,981—
19▼ 2gemini-advanced-0514ClosedGoogle1,254±352,159—
20▼ 2step-2-16k-exp-202412ClosedStepFun1,251±75,131—
21▼ 2gpt-4o-2024-08-06ClosedOpenAI1,250±347,994—
22▼ 2grok-2-2024-08-13ClosedxAI1,249±367,112—
23▼ 2yi-lightningClosed01 AI1,246±428,966—
24▼ 2gpt-4-turbo-2024-04-09ClosedOpenAI1,242±2102,190—
25▼ 2hunyuan-large-2025-02-10ClosedTencent1,242±83,862—
26▼ 2gemini-1.5-pro-001ClosedGoogle1,241±382,459—
27▼ 2claude-3-opus-20240229ClosedAnthropic1,239±2202,714—
28▼ 2gpt-4-1106-previewClosedOpenAI1,234±3103,764—
29▼ 2gpt-4o-mini-2024-07-18ClosedOpenAI1,234±368,555—
30▼ 2claude-3-5-haiku-20241022ClosedAnthropic1,232±427,327—
31▼ 2qwen2.5-plus-1127ClosedAlibaba1,231±710,729—
32▼ 2gpt-4-0125-previewClosedOpenAI1,231±397,099—
33▼ 2gemini-1.5-flash-002ClosedGoogle1,229±337,012—
34▼ 2hunyuan-standard-2025-02-10ClosedTencent1,228±94,022—
35▼ 2grok-2-mini-2024-08-13ClosedxAI1,224±355,471—
36▼ 2amazon-nova-pro-v1.0ClosedAmazon1,209±421,272—
37▼ 2reka-core-20240904ClosedReka AI1,207±67,941—
38▼ 2gemini-1.5-flash-001ClosedGoogle1,204±365,672—
39▼ 2gpt-4-0314ClosedOpenAI1,200±355,977—
40▼ 2claude-3-sonnet-20240229ClosedAnthropic1,197±2113,091—
41▼ 2reka-flash-20240904ClosedReka AI1,192±68,140—
42▼ 2glm-4-0520ClosedZhipu AI1,188±410,215—
43▼ 2gpt-4-0613ClosedOpenAI1,185±391,634—
44▼ 2amazon-nova-lite-v1.0ClosedAmazon1,180±519,290—
45▼ 2gemini-1.5-flash-8b-001ClosedGoogle1,179±437,702—
46▼ 2claude-3-haiku-20240307ClosedAnthropic1,178±3122,346—
47▼ 2amazon-nova-micro-v1.0ClosedAmazon1,163±419,356—
48▼ 2mistral-large-2402ClosedMistral1,158±364,932—
49▼ 2hunyuan-standard-256kClosedTencent1,153±102,903—
50▼ 2reka-flash-21b-20240226-onlineClosedReka AI1,151±516,022—

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 “—”.