Skip to content
IFAI

Model rankings

1388 models · 5 boards · 2 sources

‹
Week of July 27, 2026
September 7, 2026144 changesAugust 31, 202620 changesAugust 24, 202615 changesAugust 17, 202633 changesAugust 10, 2026116 changesAugust 3, 202613 changesJuly 27, 202639 changesJuly 20, 202618 changesJuly 13, 202611 changesJuly 6, 202626 changesJune 29, 202641 changesJune 22, 202612 changes
›
Artificial AnalysisText-to-image arenaArenaText-to-image arena

Text-to-image arena · Anonymous head-to-head votes · Published July 10, 2026

Best images
gpt-image-2 (medium)OpenAI1,385Score
Best open model
ideogram-4.0-qualityIdeogram1,207Score
Best from China
seedream-5.0-proByteDance1,231Score
Most voted
gemini-2.5-flash-image-preview (nano-banana)Google807,662Votes
Top three by Score
1gpt-image-2 (medium)OpenAI1,385
2reve-2.1Reve1,302
3muse-imageMeta1,280
Top three open models
1ideogram-4.0-qualityIdeogram · #13 overall1,207
2hunyuan-image-3.0Tencent · #26 overall1,151
3flux-2-devBlack Forest Labs · #28 overall1,148
This week's changes
Claude Opus 5 (Adaptive Reasoning, Max Effort)Intelligence · #1New
claude-opus-5-maxCoding · #1New
Reve 2.1Image · #2New
MiniMax H3Video · #2New
qwen3.8-maxCoding · #4New
AllOpenClosed
All regionsUSChina
One row per modelAll variants
SortScoreVotes
RankChangeModelScore95% intervalVotes
1—gpt-image-2 (medium)ClosedOpenAI1,385±560,382
2—reve-2.1ClosedReve1,302±122,432
3—muse-imageClosedMeta1,280±88,384
4—reve-2.0ClosedReve1,271±613,675
5—gemini-3.1-flash-image (nano-banana-2) [web-search]ClosedGoogle1,261±718,502
6—mai-image-2.5ClosedMicrosoft AI1,257±534,416
7—gemini-3.1-flash-lite-image (nano-banana-2-lite)ClosedGoogle1,250±89,392
8—gemini-3-pro-image-2k (nano-banana-pro)ClosedGoogle1,245±3129,467
9—gpt-image-1.5-high-fidelityClosedOpenAI1,240±3133,306
10—gemini-3-pro-image-preview (nano-banana-pro)ClosedGoogle1,232±582,565
11—seedream-5.0-proClosedByteDance1,231±112,865
12—grok-imagine-image-qualityClosedxAI1,229±439,467
13—ideogram-4.0-qualityOpenIdeogram1,207±618,547
14—qwen-image-2.0-pro-2026-06-22ClosedAlibaba1,193±78,419
15—uni-1.1-maxClosedLuma AI1,188±613,458
16—mai-image-2ClosedMicrosoft AI1,183±549,043
17—uni-1.1ClosedLuma AI1,177±520,320
18—grok-imagine-imageClosedxAI1,173±3205,635
19—recraft-v4.1-utility-proClosedRecraft1,169±112,510
20—flux-2-maxClosedBlack Forest Labs1,162±4117,111
21—grok-imagine-image-proClosedxAI1,161±493,439
22—flux-2-flexClosedBlack Forest Labs1,156±3148,925
23—flux-2-proClosedBlack Forest Labs1,155±3166,182
24—reve-v1.5ClosedReve1,154±431,694
25—gemini-2.5-flash-image-preview (nano-banana)ClosedGoogle1,151±3807,662
26—hunyuan-image-3.0OpenTencent1,151±3172,769
27—imagen-ultra-4.0-generate-001ClosedGoogle1,148±4388,009
28—flux-2-devOpenBlack Forest Labs1,148±562,500
29—seedream-4.5ClosedByteDance1,147±3220,462
30—seedream-4-2kClosedByteDance1,141±712,605
31—wan2.6-t2iClosedAlibaba1,134±3171,024
32—seedream-5.0-liteClosedByteDance1,132±487,149
33—recraft-v4.1-proClosedRecraft1,130±112,674
34—imagen-4.0-generate-001ClosedGoogle1,129±3531,357
35—qwen-image-2512OpenAlibaba · 1 more variant1,128±483,631
36—krea-2-mediumClosedKrea1,120±617,491
37—hidream-o1-imageOpenHiDream1,118±524,852
38—seedream-4-falClosedByteDance1,117±711,851
39—wan2.5-t2i-previewClosedAlibaba1,117±3219,089
40—gpt-image-1ClosedOpenAI1,115±3264,724
41—krea-2-turboOpenKrea1,113±79,143
42—seedream-4-high-res-falClosedByteDance1,113±3174,974
43—recraft-v4ClosedRecraft1,113±489,987
44—gpt-image-1-miniClosedOpenAI1,109±3163,465
45—wan2.7-image-proClosedAlibaba1,102±528,673
46—krea-2-largeClosedKrea1,100±617,210
47—wan2.7-imageClosedAlibaba1,099±529,016
48—mai-image-1ClosedMicrosoft AI1,093±494,383
49—seedream-3ClosedByteDance1,082±536,846
50—z-image-turboOpenAlibaba1,082±619,864

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 “—”.