One prompt sent to 90+ AI image models, shown side by side with cost and time
This page shows what 90+ text-to-image models make from the same prompt, side by side, with the token cost and generation time of each. The models are the ones available in the Creator Studio: GPT Image, Grok Imagine, Flux, Ideogram, Seedream, Midjourney-style models, SDXL, Pony and more, including style recipes and layout models.
The demo set uses a few fixed test prompts, each chosen to expose a different weakness: a photoreal portrait, lettering on a sign, a storybook illustration, and a counting and layout test. Switch between prompts, filter by provider, sort by cost or speed, and search by model name. Browsing the demo spends none of your tokens.
When you have a shortlist, tick the models you want and run your own prompt across them, with an optional negative prompt and aspect ratio. You see the estimated token total before you confirm, every output is billed at the listed cost, and each result is saved to your gallery. It is a quick way to decide which model to use before you start a project.
Open AI Image Model Comparison on Reality Fabricator
What you can do
- 90+ text-to-image models compared on the same prompts, with token cost and generation time
- Test prompts for a photoreal portrait, sign lettering, storybook illustration, and counting layout
- Filter by provider, sort by recommended, cheapest or fastest, and search by model name
- Pick any set of models and run your own prompt across them in one go
- Negative prompt and aspect ratio options, applied to every model that supports them
- Spend estimate shown before you confirm; every output saved to your gallery
How it works
- Open the demo tab and pick a test prompt to see what every model made from it.
- Filter by provider or sort by cost and speed to narrow the field.
- Tick the models you want to try and switch to the try-it-yourself tab.
- Enter your prompt, confirm the token estimate, and compare the results as they finish.
Frequently asked questions
Does browsing the comparison cost tokens?
No. The demo outputs are pre-generated and viewing them spends nothing. Tokens are only charged when you run your own prompt, and the page shows an estimate of the total across your picked models before you confirm.
Which image models are included?
Every text-to-image model in the Creator Studio, 90+ in total, including GPT Image, Grok Imagine, Flux, Ideogram, Seedream, Midjourney-style models, SDXL and Pony, plus the style recipes and layout models.
Can I run the same prompt on several models at once?
Yes. Pick any number of models from the demo, enter your prompt with an optional negative prompt and aspect ratio, and the page runs them together, showing each result with its cost and time.
Do I need an account to run my own prompt?
Yes. You need to log in to run a prompt, because each output is billed at the listed token cost and saved to your gallery. The demo is open to everyone without an account.
What do the test prompts check?
Each prompt targets a different weakness: a photoreal portrait for faces and skin, lettering on a sign for text rendering, a storybook illustration for style, and a counting and layout test for prompt adherence.