Build and publish in Studio · Explanation
Build and publish in Studio
Studio turns an API or an MCP server into an app with cards that people use inside Claude, ChatGPT and other AI tools.
Studio is where a workspace makes its own apps. You point it at your product, choose what the app can do, shape how its results look as cards, and try it in your own AI. Studio is in every plan, Free included; making an app public for everyone needs a paid plan.
The guides
- Studio quickstart: from your website, API document or MCP server to an app in your own AI.
- Test an app before you share it: the try page and tester links.
- Publish an app: workspace only or public, what public needs, and what outsiders get.
- Use your own domain: serve an app on a subdomain you own.
- Prepare for the Claude and ChatGPT directories: the directory kit; you submit.
What you can build from
- Your website, if it publishes an API description. Studio looks for it.
- An OpenAPI document describing your API. Studio reads the operations and lets you choose which ones become actions.
- An MCP server you already run. Studio wraps it as an app with cards.
If your product shares how to talk to it, you do not need to write code.
Plans
Studio, private apps and testing are in every plan, Free included. Making an app public, your own domain, the directory kit and payments for your app come with the Launch and Team plans. Owners and team leads add and publish a workspace's apps.
Open Studio in your workspace: app.ahel.ai/app/studio. The Studio page on ahel.ai shows what it makes.
Last updated 5 October 2026