Skip to content
IFAI

Model rankings

1092 models · 5 boards · 2 sources

Artificial AnalysisText-to-image arenaArenaText-to-image arena

Text-to-image arena · Anonymous head-to-head votes · Updated March 25, 2026

Best images
GPT Image 1.5 (high)OpenAI1,266Score
Best open model
FLUX.2 [dev] TurboFal1,165Score
Best from China
Seedream 4.0ByteDance Seed1,186Score
Top three by Score
3Riverflow 2.0Sourceful1,255
Top three open models
1FLUX.2 [dev] TurboFal · #14 overall1,165
2Qwen Image Max 2512Alibaba · #16 overall1,151
3FLUX.2 [dev]Black Forest Labs · #18 overall1,149
This week's changes
MiniMax-M2.7Intelligence · #9New
MiMo-V2-ProIntelligence · #10New
GPT-5.4 mini (xhigh)Intelligence · #14New
GLM-5-TurboIntelligence · #17New
GPT-5.4 nano (xhigh)Intelligence · #22New
RankChangeModelScore95% interval
1—Qwen Image Max 2512OpenAlibaba1,151±12
2—HunyuanImage 3.0 (Fal)OpenTencent1,121±11
3—Z-Image TurboOpenAlibaba1,115±10
4—HunyuanImage 3.0 Instruct (Fal)OpenTencent1,104±9
5—SRPOOpenTencent1,067±11
6—HunyuanImage 2.1OpenTencent1,066±10
7—Qwen ImageOpenAlibaba1,059±11
8—GLM-ImageOpenZ AI1,049±10
9▲ 2HiDream-I1-DevOpenHiDream1,047±14
10—HiDream-I1-FastOpenHiDream1,046±12
11▼ 2Infinity 8BOpenByteDance1,043±13
12—LongCat ImageOpenMeituan1,035±13
13—BagelOpenByteDance912±14
14—SDXL LightningOpenByteDance893±6
15—Janus ProOpenDeepSeek703±13

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 “—”.