> ## Documentation Index
> Fetch the complete documentation index at: https://docs.r3al.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Chat and API access

> Use the same connected model in a separate chat page or your application.

Open **Chat & API** inside a project. This page shows the model, runtime status, demo controls when configured, a separate chat URL, and a copyable API request.

## Standalone chat

Choose **Open chat** to use `/chat/{project_id}`. Share that URL with teammates who already have access to the project. It is not a public, anonymous chatbot: team sign-in is required.

Wait for **Ready**, then send a message. Follow-up messages include the current conversation as context. **New chat** clears the conversation in the page. Chat history is held in the current page and is lost on reload.

## Application access

Use the project-specific endpoint:

```text theme={null}
POST https://platform.r3al.ai/v1/projects/{project_id}/chat/completions
```

Use an account-linked API key from the same team and set `model` to the project ID. The platform forwards the request to the connected runtime; your application does not need a RunPod key.

Requests are non-streaming in this release. API calls do not start compute automatically. Start a configured demo first, or use an already-running connected runtime.

[Request and response reference](/api/inference)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.