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

# AI Models

> Compare Base and Advanced models to choose the right one for your needs

Audixa AI offers two powerful text-to-speech models to suit different needs: **Base** and **Advanced**. Each model is optimized for a different set of use cases, from general-purpose narration to highly expressive, custom voice generation.

Understanding the differences will help you choose the best model for your application.

## Base Model

The Base model is the perfect choice for a wide range of applications. It is designed for reliability, speed, and high-quality audio output across a large and diverse voice library.

<CardGroup cols={3}>
  <Card title="54 Voices" icon="users">
    Extensive library spanning various accents, genders, and styles.
  </Card>

  <Card title="High Quality" icon="waveform-lines">
    Clear, natural-sounding speech for narration and content creation.
  </Card>

  <Card title="Simple" icon="rocket">
    Easy to use with straightforward API parameters.
  </Card>
</CardGroup>

For the majority of users, the Base model provides more than enough variety and quality to build incredible audio experiences.

## Advanced Model

The Advanced model is built for users who require the highest level of customization, expressiveness, and unique vocal identity. While it has a smaller library of preset voices, its power lies in its advanced features.

<CardGroup cols={2}>
  <Card title="Voice Cloning" icon="clone">
    Create a digital replica of any voice from a short audio sample.
  </Card>

  <Card title="Custom Voices" icon="wand-magic-sparkles">
    Design unique, high-quality synthetic voices tailored to your brand.
  </Card>

  <Card title="Superior Expressiveness" icon="face-smile">
    Fine-tuned for more nuanced and emotionally resonant speech.
  </Card>

  <Card title="Greater Customization" icon="sliders">
    Detailed control over expressiveness settings like `cfg_weight` and `exaggeration`.
  </Card>
</CardGroup>

The Advanced model is ideal for enterprises, creators, and developers who need a signature voice or the flexibility of voice cloning.

## Side-by-Side Comparison

| Feature                    | Base Model                     | Advanced Model                 |
| -------------------------- | ------------------------------ | ------------------------------ |
| **Preset Voices**          | ✅ 54 Voices                    | ⚪️ Fewer (Focus on custom)     |
| **Voice Cloning**          | ❌ Not Supported                | ✅ Supported                    |
| **Custom Voices**          | ❌ Not Supported                | ✅ Supported                    |
| **Expressiveness Control** | ❌ Not Supported                | ✅ Supported                    |
| **Quality**                | High Quality & Natural         | Highest Quality & Nuanced      |
| **Best For**               | General applications           | Branding, custom experiences   |
| **API Parameter**          | `"base"`                       | `"advanced"`                   |
| **Cost**                   | [See Pricing](/guides/pricing) | [See Pricing](/guides/pricing) |

## Specifying a Model in Your API Request

To use a specific model, set the `model` parameter in your API request:

<Tabs>
  <Tab title="Base Model">
    ```json theme={null}
    {
      "text": "Hello, this is a voice from the Base model.",
      "voice_id": "am_ethan",
      "model": "base"
    }
    ```
  </Tab>

  <Tab title="Advanced Model">
    ```json theme={null}
    {
      "text": "Hello, this is a voice from the Advanced model.",
      "voice_id": "am_ethan",
      "model": "advanced",
      "cfg_weight": 3.5,
      "exaggeration": 0.7
    }
    ```
  </Tab>
</Tabs>

### SDK Examples

<CodeGroup>
  ```python Python theme={null}
  import audixa

  audixa.set_api_key("YOUR_API_KEY")

  # Using Base model
  audio_url = audixa.tts_and_wait(
      "Hello from the Base model.",
      voice_id="am_ethan",
      model="base",
  )

  # Using Advanced model with expressiveness settings
  audio_url = audixa.tts_and_wait(
      "Hello from the Advanced model!",
      voice_id="am_ethan",
      model="advanced",
      cfg_weight=3.5,
      exaggeration=0.7,
  )
  ```

  ```typescript TypeScript theme={null}
  import Audixa from 'audixa';

  const audixa = new Audixa('YOUR_API_KEY');

  // Using Base model
  const baseAudio = await audixa.generateTTS({
    text: 'Hello from the Base model.',
    voice_id: 'am_ethan',
    model: 'base'
  });

  // Using Advanced model with expressiveness settings
  const advancedAudio = await audixa.generateTTS({
    text: 'Hello from the Advanced model!',
    voice_id: 'am_ethan',
    model: 'advanced',
    cfg_weight: 3.5,
    exaggeration: 0.7
  });
  ```

  ```javascript JavaScript theme={null}
  import Audixa from 'audixa';

  const audixa = new Audixa('YOUR_API_KEY');

  // Using Base model
  const baseAudio = await audixa.generateTTS({
    text: 'Hello from the Base model.',
    voice_id: 'am_ethan',
    model: 'base'
  });

  // Using Advanced model with expressiveness settings
  const advancedAudio = await audixa.generateTTS({
    text: 'Hello from the Advanced model!',
    voice_id: 'am_ethan',
    model: 'advanced',
    cfg_weight: 3.5,
    exaggeration: 0.7
  });
  ```
</CodeGroup>

## Which Model Should I Use?

<CardGroup cols={2}>
  <Card title="Choose Base If..." icon="check">
    * You need a wide variety of preset voices
    * Cost efficiency is important
    * Standard narration and content creation
    * Quick implementation without customization
  </Card>

  <Card title="Choose Advanced If..." icon="star">
    * You need voice cloning capabilities
    * Creating a unique brand voice
    * Emotional and expressive content
    * Maximum control over output
  </Card>
</CardGroup>