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

# BYO Model Setup

> Configure Pastures to use your own AI model via any OpenAI-compatible endpoint.

Pastures works with any AI provider that implements the OpenAI chat completions API. This includes OpenAI, Anthropic, Ollama, Azure OpenAI, and any custom endpoint.

## Supported providers

| Provider     | API Base URL                          | Model example              | Auth              |
| ------------ | ------------------------------------- | -------------------------- | ----------------- |
| OpenAI       | `https://api.openai.com/v1`           | `gpt-4o`                   | API key           |
| Anthropic    | `https://api.anthropic.com`           | `claude-sonnet-4-20250514` | API key           |
| Ollama       | `http://localhost:11434/v1`           | `llama3.2`                 | None              |
| Azure OpenAI | `https://{resource}.openai.azure.com` | Your deployment name       | API key           |
| Custom       | Any OpenAI-compatible endpoint        | Provider-specific          | Provider-specific |

## Configure in Pastures

Navigate to **Pastures → Settings** and enter:

| Field        | Value                                            |
| ------------ | ------------------------------------------------ |
| AI Provider  | Your provider (OpenAI, Anthropic, Ollama, etc.)  |
| API Base URL | The provider's API endpoint                      |
| API Key      | Your provider's API key (leave blank for Ollama) |
| Model        | The model identifier                             |

## Ollama local setup

Ollama lets you run models locally with no API key and no data leaving your network.

<Steps>
  <Step title="Install Ollama">
    ```bash theme={null}
    brew install ollama
    ```

    Or download from [ollama.com](https://ollama.com).
  </Step>

  <Step title="Pull a model">
    ```bash theme={null}
    ollama pull llama3.2
    ```

    This downloads the model weights (\~4 GB for llama3.2).
  </Step>

  <Step title="Start Ollama">
    ```bash theme={null}
    ollama serve
    ```

    Ollama serves an OpenAI-compatible API at `http://localhost:11434/v1`.
  </Step>

  <Step title="Configure in Pastures">
    In **Pastures → Settings**:

    | Field        | Value                       |
    | ------------ | --------------------------- |
    | AI Provider  | Ollama                      |
    | API Base URL | `http://localhost:11434/v1` |
    | API Key      | *(leave blank)*             |
    | Model        | `llama3.2`                  |
  </Step>
</Steps>

<Tip>
  For Rancher running in a VM or container, replace `localhost` with the host machine's IP address.
</Tip>

## Limitations of BYO mode

<Info>
  BYO mode uses your model's general knowledge for diagnostics. The following features are **only available with Rancher Oracle**:

  * **Source citations** — GitHub issue links in diagnostic responses
  * **Fleet correlation** — Cross-cluster pattern matching
  * **Fix verification** — Post-fix monitoring

  See [Oracle Setup](/configuration/oracle-setup) to enable these features.
</Info>