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

# Custom Providers

> How to add and use custom LLM providers in Asserto AI

## Introduction

Asserto AI supports custom LLM providers, it uses  [Ollama-style chat completions API](https://github.com/ollama/ollama/blob/main/docs/api.md#generate-a-chat-completion).

### Chat Completions Endpoint

Example request sent by Asserto AI

Request Body:

```json theme={null}
{
  "model": "your-model-name",
  "format": "json",
  "messages": [
    {
      "role": "system",
      "content": "You are a helpful assistant."
    },
    {
      "role": "user",
      "content": "Hello, how are you?"
    }
  ],
  "stream": false
}
```

## Adding a Custom Provider

1. **Navigate to Providers**: Select *Providers* from the sidebar.
2. **Add New Provider**: Click *Add New Provider*.
3. **Configure Details**:

* Model: i.e. `llama3.2:1b`
* Base URL: Your API endpoint (e.g., `http://localhost:11434/api/generate`, `https://api.yourprovider.com/api/chat`)
* Config: You can add custom headers and query parameters to be included in all requests.

## Usage

Once configured, custom providers appear in the model selection dropdown. It can be used both in the playground and on test-runner.

<Check>Both JSON and plain text responses are supported.</Check>

<Warning>As of today, streaming and tools are not supported.</Warning>