Skip to content
IFAI

Model rankings

656 models · 3 boards · 2 sources

Artificial AnalysisIntelligence IndexArenaText arena

Text arena · Anonymous head-to-head votes · Published October 7, 2025

Most preferred
gemini-2.5-proGoogle1,452Score
Best open model
glm-4.6Z.ai1,421Score
Best from China
qwen3-max-previewAlibaba1,434Score
Most voted
claude-3-opus-20240229Anthropic196,368Votes
Top three by Score
1gemini-2.5-proGoogle1,452
Top three open models
1glm-4.6Z.ai · #13 overall1,421
2qwen3-235b-a22b-instruct-2507Alibaba · #16 overall1,419
3deepseek-v3.2-exp-thinkingDeepSeek · #17 overall1,418
This week's changes
DeepSeek V3.2 Exp (Reasoning)Intelligence · #16New
GLM-4.6 (Reasoning)Intelligence · #18New
Apriel-v1.5-15B-ThinkerIntelligence · #27New
Qwen3 4B 2507 InstructIntelligence · #88New
Granite 4.0 H SmallIntelligence · #122New
RankChangeModelScore95% intervalVotesPrice $/M
1—qwen3-max-previewClosedAlibaba1,434±616,632—
2▲ 1qwen3-max-2025-09-23ClosedAlibaba1,425±85,296—
3▼ 1glm-4.6OpenZ.ai1,421±112,935—
4—qwen3-235b-a22b-instruct-2507OpenAlibaba1,419±527,832—
5—deepseek-v3.2-exp-thinkingOpenDeepSeek · 1 more variant1,418±112,861—
6—qwen3-vl-235b-a22b-instructOpenAlibaba1,418±94,794—
7—deepseek-r1-0528OpenDeepSeek · 1 more variant1,417±619,488—
8—kimi-k2-0905-previewOpenMoonshot AI · 1 more variant1,416±710,933—
9—deepseek-v3.1-thinkingOpenDeepSeek · 1 more variant1,415±712,288—
10—deepseek-v3.1-terminusOpenDeepSeek · 1 more variant1,415±103,840—
11—glm-4.5OpenZ.ai1,407±521,427—
12—qwen3-next-80b-a3b-instructOpenAlibaba1,403±711,173—
13—longcat-flash-chatOpenMeituan1,400±710,584—
14—qwen3-235b-a22b-no-thinkingOpenAlibaba1,397±538,800—
15—qwen3-235b-a22b-thinking-2507OpenAlibaba1,397±69,483—
16—deepseek-v3-0324OpenDeepSeek · 1 more variant1,391±443,142—
17—qwen3-vl-235b-a22b-thinkingOpenAlibaba1,387±94,504—
18▲ 1qwen3-30b-a3b-instruct-2507OpenAlibaba1,385±520,396—
19▼ 1hunyuan-t1-20250711ClosedTencent1,384±94,912—
20—qwen3-coder-480b-a35b-instructOpenAlibaba1,382±522,358—
21—hunyuan-turbos-20250416ClosedTencent · 1 more variant1,379±611,218—
22—qwen3-235b-a22bOpenAlibaba1,372±526,355—
23▲ 1qwen2.5-maxClosedAlibaba1,371±433,673—
24▼ 1qwen3-next-80b-a3b-thinkingOpenAlibaba1,370±710,017—
25—glm-4.5-airOpenZ.ai1,369±520,529—
26—minimax-m1OpenMiniMax1,367±530,720—
27Newling-flash-2.0OpenAnt Group1,348±103,374—
28▼ 1glm-4.5vOpenZ.ai1,347±93,925—
29—qwen3-32bOpenAlibaba1,343±103,943—
30▼ 2step-3OpenStepFun1,343±85,724—
31▼ 1qwen-plus-0125ClosedAlibaba1,343±85,861—
32▼ 1glm-4-plus-0111ClosedZhipu · 1 more variant1,342±85,806—
33▼ 1hunyuan-turbo-0110ClosedTencent1,337±112,322—
34▼ 1qwq-32bOpenAlibaba · 1 more variant1,331±425,707—
35▼ 1step-2-16k-exp-202412ClosedStepFun1,331±84,895—
36▼ 1yi-lightningClosed01 AI1,327±527,624—
37▼ 1qwen3-30b-a3bOpenAlibaba1,324±526,618—
38▼ 1hunyuan-large-2025-02-10ClosedTencent1,323±103,760—
39▼ 1deepseek-v2.5-1210OpenDeepSeek · 1 more variant1,320±86,877—
40▼ 1step-1o-turbo-202506ClosedStepFun1,319±79,749—
41Newring-flash-2.0OpenAnt Group1,318±103,418—
42▼ 2qwen-max-0919OpenAlibaba1,316±616,598—
43▼ 2qwen2.5-plus-1127ClosedAlibaba1,312±610,252—
44▼ 2hunyuan-standard-2025-02-10ClosedTencent1,308±103,920—
45▼ 2qwen2.5-72b-instructOpenAlibaba1,300±439,632—
46▼ 2hunyuan-large-visionClosedTencent1,292±95,663—
47▼ 2glm-4-0520ClosedZhipu AI1,272±79,857—
48▼ 2qwen2.5-coder-32b-instructOpenAlibaba1,268±85,452—
49▼ 2deepseek-coder-v2OpenDeepSeek1,262±615,242—
50▼ 2qwen2-72b-instructOpenAlibaba1,261±537,688—

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 “—”.