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

# SillyTavern

> Connect SillyTavern to Kyara Intelligence as a custom OpenAI-compatible source. Set the base URL, API key, and model in a few minutes.

SillyTavern connects to Kyara as a custom OpenAI-compatible chat source. You'll enter the Kyara base URL, paste your API key, and type a model ID by hand. Have two things ready before you begin:

* Your API key from the [dashboard](https://kyara-intelligence.com/dashboard).
* A model ID from the [catalog](https://kyara-intelligence.com/models), for example `deepseek/deepseek-v4-flash`.

<Info>
  **Quick reference**

  | Setting           | Value                                   |
  | ----------------- | --------------------------------------- |
  | **Base URL**      | `https://api.kyara-intelligence.com/v1` |
  | **Model example** | `deepseek/deepseek-v4-flash`            |
</Info>

<Warning>
  The base URL must end in `/v1`. Do **not** add `/chat/completions`.
  SillyTavern appends that path for you, so
  `https://api.kyara-intelligence.com/v1/chat/completions` will fail.
</Warning>

## Set it up

<Steps>
  <Step title="Open API Connections">
    Click the **plug icon** in the top bar to open the **API Connections**
    panel.
  </Step>

  <Step title="Choose Chat Completion">
    Set the **API** dropdown to `Chat Completion`.
  </Step>

  <Step title="Choose a custom source">
    Set **Chat Completion Source** to `Custom (OpenAI-compatible)`.
  </Step>

  <Step title="Enter the base URL">
    In **Custom Endpoint (Base URL)**, paste:

    ```text theme={null}
    https://api.kyara-intelligence.com/v1
    ```
  </Step>

  <Step title="Paste your API key">
    In **Custom API Key**, paste your key from the
    [dashboard](https://kyara-intelligence.com/dashboard), then click the save
    button next to the field so it sticks.
  </Step>

  <Step title="Enter a model ID">
    In the **Model** field, type a model ID from the
    [catalog](https://kyara-intelligence.com/models), for example:

    ```text theme={null}
    deepseek/deepseek-v4-flash
    ```
  </Step>

  <Step title="Connect and test">
    Click **Connect**, then **Test Message**. A **✨ Valid ✨** confirmation
    means you're ready to chat.
  </Step>
</Steps>

## Troubleshooting

<AccordionGroup>
  <Accordion title="It won't connect">
    Check these in order:

    1. The base URL ends in `/v1` and nothing more.
    2. The key was actually saved. Re-paste it and click the save button again.
    3. The model ID has no typos. Copy it straight from the
       [catalog](https://kyara-intelligence.com/models) instead of typing it.
  </Accordion>

  <Accordion title="Which model should I use?">
    Start with `deepseek/deepseek-v4-flash` for fast, low-cost replies while you
    get set up. Once it's working, browse the
    [catalog](https://kyara-intelligence.com/models) for larger-context,
    character-tuned models that better fit your style.
  </Accordion>

  <Accordion title="A feature isn't working">
    Kyara is text-only. Tool calling, image input, and other extras are not
    available. See [unsupported features](/reference/unsupported-features) for
    the full list.
  </Accordion>
</AccordionGroup>

## Check your key first

Before opening SillyTavern, confirm your key works by sending a request from the terminal. A JSON response means your credentials are valid.

```bash theme={null}
curl https://api.kyara-intelligence.com/v1/chat/completions \
  -H "Authorization: Bearer $KYARA_INTELLIGENCE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek/deepseek-v4-flash",
    "messages": [
      { "role": "user", "content": "ping" }
    ]
  }'
```