The lineup
10 image models, one prompt each.
Every model gets the identical prompt at the same aspect ratio, once, and whatever comes back first is what gets published. The percentages below are an automated check of whether the candy’s names were printed correctly — applied the same way to every model. They are not a ranking, and the site does not publish one.
| Model | Weights | In the lineup since | Japanese name | English name | Rendered |
|---|---|---|---|---|---|
| GPT Image 2.5 SunburstOpenAI | Closed | 100%13/13 | 100%13/13 | 13/13 | |
| Nano Banana ProGoogle | Closed | 100%13/13 | 85%11/13 | 13/13 | |
| Nano Banana 2Google | Closed | 100%13/13 | 100%13/13 | 13/13 | |
| Seedream 5.0 ProByteDance | Closed | 100%13/13 | 100%13/13 | 13/13 | |
| Ideogram V4Ideogram | Closed | 77%10/13 | 69%9/13 | 13/13 | |
| Grok Imagine 2.0xAI | Closed | 92%12/13 | 100%13/13 | 13/13 | |
| Qwen Image 3Alibaba | Open | 85%11/13 | 100%13/13 | 13/13 | |
| FLUX 2 ProBlack Forest Labs | Closed | 69%9/13 | 92%12/13 | 13/13 | |
| Z-Image TurboAlibaba | Open | 46%6/13 | 69%9/13 | 13/13 | |
| MAI Image 2.5 ProMicrosoft · paused | Closed | 100%2/2 | 0%0/2 | 2/2 | |
| HunyuanImage 3Tencent | Open | 54%7/13 | 100%13/13 | 13/13 |
Sample sizes are small and shown next to every percentage on purpose. A model that has rendered a handful of candies has not been measured in any serious sense. How the check works.