Skip to content
IFAI

Model rankings

1160 models · 5 boards · 2 sources

‹
Week of April 20, 2026
June 1, 202622 changesMay 25, 202617 changesMay 18, 202616 changesMay 11, 202665 changesMay 4, 202616 changesApril 27, 202624 changesApril 20, 202627 changesApril 13, 202637 changesApril 6, 202641 changesMarch 30, 2026230 changesMarch 23, 202626 changesMarch 16, 202622 changes
›
Artificial AnalysisText-to-image arenaArenaText-to-image arena

Text-to-image arena · Anonymous head-to-head votes · Updated April 26, 2026

Best images
GPT Image 2 (high)OpenAI1,331Score
Best open model
FLUX.2 [dev] TurboFal1,170Score
Best from China
Seedream 4.0ByteDance Seed1,203Score
Top three by Score
1GPT Image 2 (high)OpenAI1,331
2GPT Image 1.5 (high)OpenAI1,271
3Nano Banana 2 (Gemini 3.1 Flash Image Preview)Google1,263
Top three open models
1FLUX.2 [dev] TurboFal · #16 overall1,170
2Qwen Image Max 2512Alibaba · #18 overall1,162
3FLUX.2 [dev]Black Forest Labs · #20 overall1,160
This week's changes
GPT-5.5 (xhigh)Intelligence · #1New
GPT Image 2 (high)Image · #1New
Realtime TTS 1.5 MaxSpeech · #1New
Claude Opus 4.7 (Adaptive Reasoning, Max Effort)Intelligence · #2New
kimi-k2.6Coding · #4New
AllOpenClosed
All regionsUSChina
One row per modelAll variants
RankChangeModelScore95% interval
1—Sana Sprint 1.6BOpenNVIDIA941±10

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 “—”.