LLM as a Service API Authentication
How to authenticate LLM as a Service API requests: one Bearer API key plus a connectionId. MindCloud handles LLM as a Service's custom auth scheme for you.
Every LLM as a Service API request through MindCloud needs two things: a MindCloud API Key and a connectionId for your LLM as a Service account. This page covers both, plus exactly how to send them, so you do not need to visit any other page to get authenticated.
Steps to authenticate
- Create a MindCloud account at app.mindcloud.co if you do not already have one.
- Generate a MindCloud API Key in API Keys. This is the Bearer token you send on every request. Keep it on your server.
- Create a LLM as a Service connection in Connections. LLM as a Service uses its own authentication scheme. You provide whatever it requires when you create the connection — MindCloud handles the provider-specific details for you.
- Copy the connection’s
connectionId— you will send it on every request alongside your MindCloud API Key.
How requests are authenticated
Authentication has two layers. Your MindCloud API Key authenticates the request to MindCloud, sent as a Bearer token in the Authorization header. The connectionId selects the connected LLM as a Service account that should run the action. MindCloud stores and refreshes the LLM as a Service credentials behind that connection, so you never send provider tokens with your requests.
Pass connectionId in the query string for GET and DELETE actions, and in the JSON body for POST, PUT, and PATCH actions.
For example, calling with an LLM as a Service connection looks like this:
The Authorization: Bearer <your MindCloud API Key> header authenticates you to MindCloud; the connectionId tells MindCloud which LLM as a Service account to run the action against. Missing or invalid Authorization returns 401, and a missing or mismatched connectionId returns 400.
Prepare your LLM as a Service project
You need an LLM as a Service account, a project, and at least one configured LLM service in the control panel.
- Sign in or create your account
Open the LLM as a Service control panel, register if needed, and confirm your email address.
Control Panel - Create a project
After your first login, create a project when the control panel prompts you to do so.
- Open LLMServices
In the control panel, open the LLMServices area from the right-side menu.
- Add at least one LLM service
Create a service endpoint for the model provider you want this project to use.
- Save the provider API key and test the service
Use the provider API-key flow in the LLM service editor, save the vendor key there, and run the built-in test call before continuing.
- Open the Integration page
Go to the Integration page and locate the credentials area for your project.
Integration API - Copy your Project ID
Copy the Project ID from the Integration page and use it in the MindCloud connection.
Keep your key safe
Keep your MindCloud API Key on your server; do not ship it in browser code, mobile apps, public repositories, or logs. Anyone with the key can call the Universal API as your MindCloud account, across every connection you have created.