For platforms

What we build for a platform

• A library of styles. Our own set of unique style prompts, each evoking one particular look. Add one to your prompt, and your idea comes out in that style. Every style is rendered on your model and judged by eye before it’s published.
• A home on Midlibrary. A hub of its own and its own place in the library, searchable by look, color and subject.
• An editorial pool. Guides, experiments, compilations and know-hows that teach the recipes of your model: how it works and how to get the best out of it, explained plainly and richly illustrated, every picture with its prompt. Written to make people want to open your platform and try it themselves.

What we ask

Access for generating at scale (an API and tokens), and cross-marketing when the library launches: your channels and ours, on the same day.

That’s all. It’s free for the platform.

Which platforms we build libraries for

• A visual model of its own, or a distinct take on one.
• People really make work with it.
• Access we can generate with at scale.

We choose by these, never by budget.

How we work

• Nobody pays for placement. Not in the library, not in the guides.
• Nobody approves our words. If we get a fact about your platform wrong, tell us and we’ll fix it. The wording stays ours.
• First, never only. A new platform gets its own launch and about two months of quiet after it. No platform gets exclusive rights to prompts, categories or the platforms that follow.
• Readers see these rules too. Every guide that recommends a platform links here, and this page lists what each platform has given us.

Other ways to work together

Building something and want a second pair of eyes from people who make pictures with visual AI every day? We consult, test and collaborate beyond the library too. Just ask.