Image to Image

Image to image starts from a picture you already have and returns a modified version of it. You supply the source image plus an instruction, and the model edits in place: change the background, restyle the whole frame, remove or add an object, adjust colour and lighting, or clean up a photograph. The original composition stays as the anchor, which is what separates this from generating something new.

The clearer the instruction, the better the edit. Say what should change and, just as importantly, what should stay untouched. Edits compound, so several small passes usually hold likeness better than one sweeping instruction. Expect some drift in fine detail such as faces, hands and small type on every pass, and keep the untouched original so you can restart from it rather than editing an edit five times over.

The models serving this category on E2X are the editing modes of Google's Nano Banana family, including the Pro and Lite variants, together with OpenAI's GPT Image 2. They run through the same job endpoint as everything else in the catalog: post a model slug with your input object, then poll the job id. Pricing is per request from a prepaid balance, so an edit that misses costs one request to retry rather than a subscription tier.