Create from your agent or your app.

Yapper runs a hosted MCP server. Paste the connector URL into any MCP-compatible client, sign in with your Yapper account, and your agent gains a set of tools for listing models, checking credits, browsing assets, and starting generation jobs.

Any client that supports hosted, remote, or Streamable HTTP MCP servers — including Claude, ChatGPT, Cursor, Windsurf, Cline, and VS Code. Custom agents work too.

The same models available in the Yapper app, including image models like Nano Banana Pro, GPT Image 2, and Seedream, and video models like Sora 2, Veo 3.1, Kling 3.0, and Seedance 2.0 — plus upscalers and other tools.

Not for the hosted connector — you sign in with your Yapper account. API keys are available separately for building your own apps, scripts, and backend integrations against the Yapper REST API.

Everything draws from your existing Yapper credit balance. Each model has a representative starting price, and you can request an exact quote for any job with a dry run before spending.

It depends on the model and settings — images typically return in seconds, and video in under a few minutes. Jobs run asynchronously, and you can poll for status or receive a webhook on completion.

Yes. Import media from any URL, reference your own uploads for style and character consistency, and browse assets already in your Yapper library — all from the conversation.

Absolutely. Beyond the connector, the Yapper REST API lets you start processes, import and manage assets, check usage, and wire generation into any product or workflow.