Model categories

A category on E2X groups models by what they take in and what they hand back, not by who trained them. If you have a written description and want a still image, text to image is the shelf to look at. If you have a finished frame and want motion, image to video is. That framing keeps the choice practical: pick the capability you need first, then compare the models that serve it on price, speed and the quality of what they return.

Every category page lists the models the catalog is serving right now, with the per-request price attached to each one. Those prices come from the same catalog the API reads, so the number on the page is the number your account is charged. Because all six capabilities run through one endpoint, moving between categories is a change of model slug and input fields rather than a new integration, a second SDK or another authentication flow.

The six capabilities live today are text to image, image to image, text to video, image to video, reference to video, and start and end frame to video. Coverage is currently deeper on the image side than on video, and the catalog is the honest place to check that: it shows what is available at this moment rather than what is planned. Billing is a prepaid balance drawn down per request, with no monthly platform fee and no per-seat charge, so trying several categories costs only the requests you actually send.