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

# Audio

> Text-to-speech and speech-to-text

Agents can speak and listen - convert text to speech and transcribe audio.

```mermaid theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
graph LR
    subgraph "Audio Processing"
        A[👤 User] --> B[🤖 Agent]
        B -->|speak| C[🔊 Audio]
        B -->|listen| D[📝 Text]
    end
    
    classDef user fill:#6366F1,stroke:#7C90A0,color:#fff
    classDef agent fill:#8B0000,stroke:#7C90A0,color:#fff
    classDef output fill:#10B981,stroke:#7C90A0,color:#fff
    
    class A user
    class B agent
    classDef tool fill:#189AB4,stroke:#7C90A0,color:#fff
    class C,D output
    classDef agent fill:#8B0000,color:#fff
    classDef tool fill:#189AB4,color:#fff
```

## Quick Start

<Steps>
  <Step title="Simple Usage">
    ```typescript theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
    import { Agent } from 'praisonai';

    const agent = new Agent({
      instructions: 'You are a helpful assistant',
      audio: true
    });

    // Agent response as audio
    const audio = await agent.speak('Hello! How can I help you today?');
    // Returns audio buffer
    ```
  </Step>

  <Step title="With Configuration">
    ```typescript theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
    // Transcribe audio file
    const text = await agent.transcribe('./recording.mp3');
    console.log(text);
    // "This is what was said in the recording..."
    ```
  </Step>
</Steps>

***

## User Interaction Flow

```mermaid theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
sequenceDiagram
    participant User
    participant Agent
    participant Audio
    
    User->>Agent: "Read this aloud"
    Agent->>Audio: Generate speech
    Audio-->>User: 🔊 Audio playback
    
    User->>Agent: [voice recording]
    Agent->>Audio: Transcribe
    Audio-->>Agent: Text content
    Agent-->>User: "You said..."
```

***

## Configuration Levels

```typescript theme={"theme":{"light":"vitesse-light","dark":"vitesse-dark"}}
// Level 1: Bool - Enable with defaults
const agent = new Agent({
  audio: true
});

// Level 2: String - Specify voice
const agent = new Agent({
  audio: 'alloy'  // OpenAI voice name
});

// Level 3: Dict - Full options
const agent = new Agent({
  audio: {
    voice: 'nova',
    model: 'tts-1-hd',
    speed: 1.0,
    format: 'mp3'
  }
});

// Level 4: Instance - AudioAgent
import { AudioAgent } from 'praisonai';

const audio = new AudioAgent({
  provider: 'elevenlabs',
  voice: 'rachel'
});
```

***

## Audio Options

| Option   | Description                          |
| -------- | ------------------------------------ |
| `voice`  | Voice name (alloy, echo, nova, etc.) |
| `model`  | TTS model (tts-1, tts-1-hd)          |
| `speed`  | Playback speed (0.25 to 4.0)         |
| `format` | Output format (mp3, wav, opus)       |

***

## API Reference

<Card title="AudioConfig" icon="code" href="/docs/sdk/reference/typescript/classes/AudioConfig">
  Complete configuration options
</Card>

***

## Best Practices

<AccordionGroup>
  <Accordion title="Use HD for quality">
    `tts-1-hd` sounds more natural but costs more.
  </Accordion>

  <Accordion title="Match voice to content">
    Choose voices that match your content's tone and audience.
  </Accordion>

  <Accordion title="Handle long text">
    Break long text into chunks for better audio quality.
  </Accordion>
</AccordionGroup>

***

## Related

<CardGroup cols={2}>
  <Card title="Voice" icon="microphone" href="/docs/js/voice">
    Voice conversations
  </Card>

  <Card title="Realtime" icon="bolt" href="/docs/js/realtime">
    Real-time streaming
  </Card>
</CardGroup>