Skip to content
IFAI

Model rankings

298 models · 2 boards · 2 sources

Artificial AnalysisIntelligence IndexArenaText arena

Text arena · Anonymous head-to-head votes · Published April 9, 2025

Most preferred
Best open model
deepseek-v3-0324DeepSeek1,321Score
Best from China
deepseek-v3-0324DeepSeek1,321Score
Most voted
claude-3-opus-20240229Anthropic202,710Votes
Top three by Score
Top three open models
1deepseek-v3-0324DeepSeek · #4 overall1,321
2deepseek-r1DeepSeek · #6 overall1,316
3gemma-3-27b-itGoogle · #14 overall1,281
This week's changes
Llama 4 MaverickIntelligence · #11New
Llama 4 ScoutIntelligence · #22New
Gemma 3 12B InstructIntelligence · #51New
Gemma 3 4B InstructIntelligence · #71New
Gemini 1.5 Flash (Sep '24)Intelligence · #35Up 20
RankChangeModelScore95% intervalVotesPrice $/M
1—chatgpt-4o-latest-20250326ClosedOpenAI1,360±66,612—
2—gpt-4.5-preview-2025-02-27ClosedOpenAI1,356±613,443—
3—grok-3-preview-02-24ClosedxAI1,327±513,919—
4—o1-2024-12-17ClosedOpenAI1,320±427,831—
5—claude-3-7-sonnet-20250219-thinking-32kClosedAnthropic1,314±67,809—
6—o1-previewClosedOpenAI1,303±433,175—
7—claude-3-7-sonnet-20250219ClosedAnthropic1,300±613,107—
8—o3-mini-highClosedOpenAI1,287±516,889—
9▲ 1qwen2.5-maxClosedAlibaba1,286±519,995—
10▼ 1claude-3-5-sonnet-20241022ClosedAnthropic · 1 more variant1,285±364,098—
11—gemini-2.0-flash-001ClosedGoogle1,283±521,065—
12▲ 1o3-miniClosedOpenAI1,269±523,693—
13▼ 1gemini-1.5-pro-002ClosedGoogle1,268±358,651—
14—gemini-2.0-flash-lite-preview-02-05ClosedGoogle1,268±520,990—
15—hunyuan-turbos-20250226ClosedTencent1,265±112,456—
16—gpt-4o-2024-05-13ClosedOpenAI1,263±2117,768—
17—qwen-plus-0125ClosedAlibaba1,262±76,058—
18—hunyuan-turbo-0110ClosedTencent1,258±102,509—
19—glm-4-plus-0111ClosedZhipu · 1 more variant1,256±76,032—
20—o1-miniClosedOpenAI1,255±454,967—
21—gemini-advanced-0514ClosedGoogle1,254±352,136—
22—step-2-16k-exp-202412ClosedStepFun1,251±85,126—
23—gpt-4o-2024-08-06ClosedOpenAI1,250±347,992—
24—grok-2-2024-08-13ClosedxAI1,249±467,092—
25—yi-lightningClosed01 AI1,246±428,971—
26—gpt-4-turbo-2024-04-09ClosedOpenAI1,242±3102,160—
27—hunyuan-large-2025-02-10ClosedTencent1,242±93,860—
28—gemini-1.5-pro-001ClosedGoogle1,241±382,446—
29—claude-3-opus-20240229ClosedAnthropic1,239±2202,710—
30—gpt-4-1106-previewClosedOpenAI1,235±3103,753—
31—gpt-4o-mini-2024-07-18ClosedOpenAI1,234±370,818—
32—claude-3-5-haiku-20241022ClosedAnthropic1,233±432,053—
33▲ 1qwen2.5-plus-1127ClosedAlibaba1,231±610,716—
34▼ 1gpt-4-0125-previewClosedOpenAI1,231±397,077—
35—gemini-1.5-flash-002ClosedGoogle1,229±437,026—
36—hunyuan-standard-2025-02-10ClosedTencent1,229±94,014—
37—grok-2-mini-2024-08-13ClosedxAI1,224±355,449—
38—amazon-nova-pro-v1.0ClosedAmazon1,209±623,635—
39—reka-core-20240904ClosedReka AI1,207±87,938—
40—gemini-1.5-flash-001ClosedGoogle1,204±365,668—
41—gpt-4-0314ClosedOpenAI1,200±355,980—
42—claude-3-sonnet-20240229ClosedAnthropic1,197±2113,056—
43—reka-flash-20240904ClosedReka AI1,193±88,138—
44—glm-4-0520ClosedZhipu AI1,188±610,221—
45—gpt-4-0613ClosedOpenAI1,185±291,639—
46—amazon-nova-lite-v1.0ClosedAmazon1,182±420,655—
47—gemini-1.5-flash-8b-001ClosedGoogle1,179±437,697—
48—claude-3-haiku-20240307ClosedAnthropic1,178±2122,305—
49—amazon-nova-micro-v1.0ClosedAmazon1,164±520,666—
50—mistral-large-2402ClosedMistral1,158±364,915—

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 “—”.