> ## Documentation Index
> Fetch the complete documentation index at: https://mintlify.com/n4ze3m/page-assist/llms.txt
> Use this file to discover all available pages before exploring further.

# LM Studio

> Integrate LM Studio with Page Assist

LM Studio is a desktop application that lets you run large language models locally with an easy-to-use interface. Page Assist can connect to LM Studio through its OpenAI-compatible API server.

## Prerequisites

* LM Studio installed on your system
* At least one model downloaded in LM Studio
* LM Studio's local server enabled

## Installation

If you haven't installed LM Studio yet:

<Steps>
  <Step title="Download LM Studio">
    Visit [lmstudio.ai](https://lmstudio.ai) and download the installer for your operating system.
  </Step>

  <Step title="Install LM Studio">
    Run the installer and follow the installation instructions.
  </Step>

  <Step title="Download a Model">
    Open LM Studio and browse the model library. Download a model suitable for your hardware:

    * **Small (3B-7B)**: Good for 8-16GB RAM
    * **Medium (13B)**: Requires 16-32GB RAM
    * **Large (30B+)**: Requires 32GB+ RAM

    Popular choices:

    * Llama 3.2 (3B or 8B)
    * Mistral 7B
    * Phi-3 (3.8B)
  </Step>

  <Step title="Load the Model">
    In LM Studio, click on a downloaded model to load it into memory.
  </Step>
</Steps>

## Enabling the API Server

LM Studio must have its local server running for Page Assist to connect:

<Steps>
  <Step title="Open Local Server Tab">
    In LM Studio, navigate to the "Local Server" or "Developer" tab (location varies by version).
  </Step>

  <Step title="Start the Server">
    Click "Start Server". The default address is:

    ```text theme={null}
    http://localhost:1234/v1
    ```
  </Step>

  <Step title="Verify Server is Running">
    You should see a confirmation message that the server is running and listening on port 1234.
  </Step>
</Steps>

## Connecting to Page Assist

Once LM Studio's server is running, connect it to Page Assist:

<Steps>
  <Step title="Open Page Assist Settings">
    Click the Page Assist icon in your browser toolbar, then click the Settings icon.
  </Step>

  <Step title="Navigate to OpenAI Compatible API">
    Go to the "OpenAI Compatible API" tab.
  </Step>

  <Step title="Add Provider">
    Click the "Add Provider" button.
  </Step>

  <Step title="Select LM Studio">
    Choose "LMStudio" from the provider dropdown menu.
  </Step>

  <Step title="Enter API URL">
    Enter the LM Studio API URL. The default is:

    <CodeGroup>
      ```text Default theme={null}
      http://localhost:1234/v1
      ```

      ```text Custom Port theme={null}
      http://localhost:8080/v1
      ```

      ```text Remote Server theme={null}
      http://192.168.1.100:1234/v1
      ```
    </CodeGroup>
  </Step>

  <Step title="Save Configuration">
    Click "Save". Page Assist will automatically fetch available models from LM Studio.
  </Step>
</Steps>

## Model Management

### Automatic Model Detection

Page Assist automatically detects models that are currently loaded in LM Studio. The model must be loaded in LM Studio before it appears in Page Assist.

### Switching Models

To use a different model:

1. In LM Studio, stop the current model
2. Load the desired model
3. Refresh Page Assist or restart the chat
4. The new model should appear in the model selector

### Model Not Appearing

If a model doesn't appear in Page Assist:

1. Ensure the model is loaded in LM Studio (not just downloaded)
2. Verify the API server is running
3. Check the connection URL in Page Assist settings
4. Try refreshing the Page Assist interface

## Configuration Options

### Custom Server Port

To change LM Studio's server port:

1. In LM Studio's server settings, change the port number
2. Update the URL in Page Assist to match:
   ```text theme={null}
   http://localhost:[NEW_PORT]/v1
   ```

### Server Authentication

LM Studio's local server typically doesn't require authentication. If you've configured API keys:

1. In Page Assist provider settings, expand advanced options
2. Enter your API key in the authentication field
3. Save the configuration

### Custom Headers

For advanced configurations requiring custom headers:

1. In provider settings, find "Custom Headers" option
2. Add required headers as key-value pairs
3. Save your configuration

## Troubleshooting

### Connection Failed

<Steps>
  <Step title="Verify Server is Running">
    Check that LM Studio's local server shows as "Running".
  </Step>

  <Step title="Check Port Number">
    Ensure the port in Page Assist matches LM Studio's server port (default: 1234).
  </Step>

  <Step title="Test URL in Browser">
    Navigate to `http://localhost:1234/v1/models` in your browser. You should see a JSON response with model information.
  </Step>

  <Step title="Check Firewall">
    Ensure your firewall isn't blocking localhost connections on port 1234.
  </Step>

  <Step title="Restart LM Studio Server">
    Stop and restart the server in LM Studio.
  </Step>
</Steps>

### Model Not Loaded

Error messages like "No model loaded":

1. Load a model in LM Studio by clicking on it
2. Wait for the model to fully load into memory
3. Verify "Model Loaded" appears in LM Studio
4. Try sending a message again

### Slow Response Times

If responses are slow:

* **Use a smaller model**: Switch to a 3B or 7B model
* **Adjust GPU layers**: In LM Studio, increase GPU layer count if you have a GPU
* **Check system resources**: Close other applications to free up RAM
* **Use quantized models**: GGUF Q4 or Q5 quantizations are faster

### Out of Memory Errors

1. Use a smaller model
2. Close other applications
3. Reduce context length in LM Studio settings
4. Restart LM Studio to clear memory

## Performance Optimization

### GPU Acceleration

If you have a compatible GPU:

1. In LM Studio, go to settings
2. Enable GPU acceleration
3. Set GPU layer count (higher = more GPU usage)
4. Reload your model

### Context Length

Adjust context window size based on your needs:

* **Shorter contexts (2048-4096)**: Faster, less memory
* **Longer contexts (8192+)**: Slower, more memory, better for long conversations

### Batch Size

In LM Studio settings:

* Increase batch size for faster generation (requires more memory)
* Decrease if you encounter memory issues

## Multiple LM Studio Instances

You can run multiple LM Studio instances on different ports:

1. Start first instance on default port 1234
2. Start second instance on a different port (e.g., 1235)
3. Add each as a separate provider in Page Assist

## Comparison with Ollama

| Feature          | LM Studio           | Ollama                     |
| ---------------- | ------------------- | -------------------------- |
| UI               | Graphical           | Command-line               |
| Model Management | Built-in browser    | CLI commands               |
| Ease of Use      | Very Easy           | Moderate                   |
| Model Format     | GGUF                | GGUF                       |
| API              | OpenAI-compatible   | Ollama & OpenAI-compatible |
| Platform         | Windows, Mac, Linux | Windows, Mac, Linux        |

## Best Practices

1. **Keep LM Studio Updated**: Regular updates improve performance and compatibility
2. **Monitor Memory**: Watch RAM usage and choose models accordingly
3. **One Model at a Time**: Load only one model at a time to save resources
4. **Persistent Server**: Keep the API server running while using Page Assist
5. **Model Selection**: Test different models to find the best balance of quality and speed

## Next Steps

* Learn about [Configuration Settings](/configuration/settings)
* Explore [Knowledge Base features](/features/knowledge-base)
* Set up [additional providers](/providers/openai-compatible)