Skip to content
IFAI

Model rankings

769 models · 5 boards · 2 sources

‹
Week of January 19, 2026
March 2, 202663 changesFebruary 23, 202613 changesFebruary 16, 202616 changesFebruary 9, 202641 changesFebruary 2, 20266 changesJanuary 26, 2026177 changesJanuary 19, 202610 changesJanuary 12, 2026116 changesJanuary 5, 20263 changesDecember 29, 202513 changesDecember 22, 20256 changesDecember 15, 20258 changes
›
Artificial AnalysisIntelligence IndexArenaText arena

Intelligence Index · A composite of several evaluations · Updated January 14, 2026

Smartest
GPT-5.2 (xhigh)OpenAI50.6Intelligence
Best open model
GLM-4.7 (Reasoning)Z AI41.7Intelligence
Best from China
GLM-4.7 (Reasoning)Z AI41.7Intelligence
Top three by Intelligence
1GPT-5.2 (xhigh)OpenAI50.6
2Claude Opus 4.5 (Reasoning)Anthropic49.1
3Gemini 3 Pro Preview (high)Google47.9
Top three open models
1GLM-4.7 (Reasoning)Z AI · #9 overall41.7
2DeepSeek V3.2 (Reasoning)DeepSeek · #12 overall41.2
3Kimi K2 ThinkingKimi · #14 overall40.3
This week's changes
gpt-5.2-codexCoding · #17New
FLUX.2 [klein] 9BImage · #21New
FLUX.2 [klein] Base 9BImage · #43New
FLUX.2 [klein] 4BImage · #48New
GLM-ImageImage · #49New
AllOpenClosed
All regionsUSChina
One row per modelAll variants
SortIntelligenceValueSpeedLowest price
RankChangeModelIntelligencePrice $/MSpeed tok/sFirst token
1—GLM-4.7 (Reasoning)OpenZ AI41.7———
2—DeepSeek V3.2 (Reasoning)OpenDeepSeek41.2———
3—Kimi K2 ThinkingOpenKimi40.3———
4—MiniMax-M2.1OpenMiniMax39.4———
5—MiMo-V2-Flash (Reasoning)OpenXiaomi39.1———
6—MiniMax-M2OpenMiniMax35.7———
7—GLM-4.7 (Non-reasoning)OpenZ AI33.8———
8—DeepSeek V3.1 Terminus (Reasoning)OpenDeepSeek33.5———
9—Doubao Seed CodeClosedByteDance Seed33.2———
10—DeepSeek V3.2 Exp (Reasoning)OpenDeepSeek32.5———
11—Qwen3 Max ThinkingClosedAlibaba32.3———
12—GLM-4.6 (Reasoning)OpenZ AI32.2———
13—DeepSeek V3.2 (Non-reasoning)OpenDeepSeek31.9———
14—Qwen3 MaxClosedAlibaba31.0———
15—MiMo-V2-Flash (Non-reasoning)OpenXiaomi30.1———
16—GLM-4.6 (Non-reasoning)OpenZ AI29.8———
17—Qwen3 235B A22B 2507 (Reasoning)OpenAlibaba29.4———
18—DeepSeek V3.2 SpecialeOpenDeepSeek29.0———
19—ERNIE 5.0 Thinking PreviewClosedBaidu28.9———
20—Kimi K2 0905OpenKimi28.0———
21—DeepSeek V3.2 Exp (Non-reasoning)OpenDeepSeek28.0———
22—DeepSeek V3.1 Terminus (Non-reasoning)OpenDeepSeek28.0———
23—DeepSeek V3.1 (Reasoning)OpenDeepSeek27.9———
24—DeepSeek V3.1 (Non-reasoning)OpenDeepSeek27.7———
25—Qwen3 VL 235B A22B (Reasoning)OpenAlibaba27.5———
26—DeepSeek R1 0528 (May '25)OpenDeepSeek27.1———
27—GLM-4.5 (Reasoning)OpenZ AI26.8———
28—Qwen3 Next 80B A3B (Reasoning)OpenAlibaba26.5———
29—Seed-OSS-36B-InstructOpenByteDance Seed25.0———
30—Qwen3 Coder 480B A35B InstructOpenAlibaba24.8———
31—Qwen3 235B A22B 2507 InstructOpenAlibaba24.6———
32—MiniMax M1 80kOpenMiniMax24.3———
33—GLM-4.6V (Reasoning)OpenZ AI23.8———
34—GLM-4.5-AirOpenZ AI23.6———
35—Qwen3 30B A3B 2507 (Reasoning)OpenAlibaba22.8———
36—DeepSeek V3 0324OpenDeepSeek22.6———
37—Qwen3 VL 235B A22B InstructOpenAlibaba20.8———
38—Qwen3 Next 80B A3B InstructOpenAlibaba20.3———
39—Qwen3 Coder 30B A3B InstructOpenAlibaba20.2———
40—Qwen3 235B A22B (Reasoning)OpenAlibaba19.9———
41—Ling-1TOpenInclusionAI19.7———
41—Qwen3 VL 30B A3B (Reasoning)OpenAlibaba19.7———
43—Qwen3 4B 2507 (Reasoning)OpenAlibaba18.8———
44—DeepSeek R1 (Jan '25)OpenDeepSeek18.7———
45—GLM-4.6V (Non-reasoning)OpenZ AI17.4———
46—Qwen3 235B A22B (Non-reasoning)OpenAlibaba17.3———
47—Qwen3 VL 32B InstructOpenAlibaba17.1———
48—Qwen3 32B (Reasoning)OpenAlibaba17.1———
49—Qwen3 VL 8B (Reasoning)OpenAlibaba17.0———
50—DeepSeek V3 (Dec '24)OpenDeepSeek17.0———

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 “—”.