120 models. One credit balance.
Every model in the studio has its own page — what it costs, what it takes as input, the looks it is good at, and the public work people have made on it.
39 image models
FAQ
Before you pick a model.
It depends on the model class, not the media type. A fast still is 1 credit, a standard still 5 and a premium still 20. A fast clip is 60 credits, a premium clip 140 and a flagship clip like Veo 3.1 or Sora 2 is 400. A music track is 80 and an upscale is 10. Models in the free experimental pool cost nothing. Every model page states its own figure, and so does the Generate button before you commit.
Yes. The whole library is reachable from every tier, including Free — the difference between plans is how much you can render, how fast the queue moves and what licence the output carries. A free account gets daily credits, so the cheaper models can be run immediately and a premium render is a matter of saving a day of credits.
Those models run on a shared pool that costs no credits. In exchange they are slower, they queue behind other free traffic, and resolution or duration is requested rather than guaranteed. They are ideal for drafting an idea before you spend credits finishing it.
Start from what the brief needs rather than from a leaderboard: typography-heavy work suits GPT Image, photoreal portraits suit the Seedream and Nano Banana families, and anything that has to animate later is best started on a model whose output you can feed to an image-to-video model. When two look equally plausible, run them side by side on the same prompt and let the frames decide.
Edit models and image-to-video models do — their model page says "Text + a reference image" or "Text + a start frame" under Input. Everything else works from a text prompt alone.
Free-tier output carries a subtle Ideovex watermark. Every paid tier downloads clean, and Creator and Studio include a commercial usage licence.
Pick a model and run it
Free credits on signup, no card required. Swap models mid-project without swapping tools.




































