Reve

Reve is an image generation and editing platform whose model was trained with a heavy emphasis on two things creatives complain about most: following the prompt exactly, and rendering legible text inside the image.

What it does well

• Prompt adherence. Detailed compositional instructions — who is where, holding what, in which style — survive into the result.
• Typography. Posters, packaging, signage, and UI mockups come out with text you can actually read, which removes a whole round of manual correction.
• Conversational editing. Describe a change to an existing image and only that changes. You can iterate on one image across many turns instead of rerolling.
• Reference images. Supply a subject or a style reference to keep outputs consistent across a set.

How people use it

Design and marketing work where the image has to communicate something specific: campaign concepts, social assets, mockups, and editorial illustration. The editing loop is the strongest part of the product — it behaves like working with a designer rather than a slot machine.

Free credits let you evaluate it; paid plans add volume, higher resolution, and commercial usage rights. An API is available for embedding it in your own tools.