AA Text-to-Image
- What does it measure?
- A ranking of which picture people chose when two AIs drew from the same prompt, shown side by side. It works like the Arena Intelligence image ranking but uses a separate pool of votes from another organization, so reading both helps filter out one voter pool’s taste.
- Who evaluates it, and how?
- In the Image Arena of independent evaluator Artificial Analysis, users pick one of two anonymous images made from the same prompt, and the choices are scored Elo-style. Settings such as resolution and quality are ranked separately.
- Reading the score
- This is a head-to-head record, not an exam score. There is no maximum; a 100-point gap means the leader is preferred about 64% of the time. When an AI has several quality settings, the setting with the most samples is shown, and AIs with few samples can swing widely. The scale differs from Arena Intelligence, so do not compare the numbers directly. Values in this table are currently about 850–1,200.
Try it yourself
Give two AIs the same prompt and pick the better picture.
Two AIs draw from one line of text; you vote with names hidden.
Rank
Top score1261
Models tested38
Scoring basisHigher is better
Last updated2026.11.20
Model release date
Which AI makes the images people prefer most?Arena T2I
GPT Image 2.5 Sunburst🥇
GPT Image 2.5 Sunburst🥇Arena T2I
GPT Image 2.5 Sunburst🥇Which AI edits images the way people prefer most?Arena Image Edit
GPT Image 2.5 Sunburst🥇
GPT Image 2.5 Sunburst🥇Arena Image Edit
GPT Image 2.5 Sunburst🥇Which AI do people prefer most for turning photos into videos?Arena Image-to-Video
Gemini Omni 1.2 Flash🥇
Gemini Omni 1.2 Flash🥇Arena Image-to-Video
Gemini Omni 1.2 Flash🥇
Which AI makes the images people prefer most?
AA Text-to-Image
- 🥇OpenAIGPT Image 2.6 Sunburst1261
- 🥈OpenAIGPT Image 2.5 Sunburst1230
- 🥉OpenAIGPT Image 2.6 Flare1227
- #4MicrosoftMAI-Image-2.71215
- #5OpenAIGPT Image 2.5 Flare1211
- #6OpenAIGPT Image 21193
- #7GoogleNano Banana 2.11177
- #8MicrosoftMAI-Image-2.61168
- #9SpaceXSIGrok Imagine Image 2.01167
- #10MicrosoftMAI-Image-2.7 Flash1159
- #11GoogleNano Banana 21144
- #12ByteDanceSeedream 5.51134
- #13OpenAIGPT Image 1.51126
- #13MetaMuse Image1126
- #15MicrosoftMAI-Image-2.6 Flash1123
- #16GoogleNano Banana Pro1119
- #17MicrosoftMAI-Image-2.51118
- #18Black Forest LabsFLUX 3 Image1116
- #18MicrosoftMAI-Image-2.5 Pro1116
- #20GoogleNano Banana 2 Lite1110
- #21ByteDanceSeedream 5.0 Pro1104
- #22MiniMaxMiniMax H3 Image Highspeed1072
- #23MiniMaxMiniMax H3 Image1066
- #24AlibabaQwen Image 2.1 Pro1058
- #25SpaceXSIGrok Imagine Image Quality1050
- #26ByteDanceSeedream 4.01031
- #27ByteDanceSeedream 4.51030
- #28Black Forest LabsFLUX.2 Flex1028
- #29AlibabaQwen Image 2.1 Turbo1025
- #30Black Forest LabsFLUX.2 Max1022
- #31ByteDanceSeedream 5.0 Lite1020
- #32OpenAIGPT Image 11018
- #33Black Forest LabsFLUX.2 Pro1005
- #34RecraftRecraft V4.1991
- #35RecraftRecraft V4984
- #36RecraftRecraft V4.1 Flash950
- #37OpenAIGPT Image 1 Mini939
- #38Black Forest LabsFLUX.2 Klein 4B840
Half of models ≤ 1113Measured by: User preference votes
4 of the 16 AIs released in the last 60 days have no score here yet · Recraft V4.1 Flash V2, HY-World 2.1 Panorama, Riverflow V2.6 Pro and 1 more
It’s a preference vote, not a check that the request was followed. A result that looks or sounds nicer can win even if it strays.
Which requests got made depends on who voted — they may differ from your use.
Scores labelled 'Self-reported' were published by the AI maker itself, not measured by a third party.