Text-to-speech (TTS)

Octave TTS, the first text-to-speech system built on LLM intelligence. Unlike conventional TTS that merely “reads” words, Octave is a “speech-language model” that understands what words mean in context, unlocking a new level of expressiveness and nuance.

At Octave’s core it is a state-of-the-art large language model (LLM) that Hume AI trained to understand and synthesize speech. This speech-language model can predict the tune, rhythm and timbre of speech, knowing when to whisper secrets, shout triumphantly, or calmly explain a fact. This combined approach lets Octave interpret plot twists, emotional cues, and character traits within a script or prompt, then transform that understanding into lifelike speech.

Features

Using text-to-speech

The TTS API provides a RESTful interface for generating expressive speech from text. You send text to synthesize along with optional voice specifications and descriptions, and receive audio in your chosen format.

Basic speech synthesis

At its simplest, you can generate speech by sending text with an optional voice description. The description helps shape the voice’s characteristics and expression:

1
curl -X POST "https://api.hume.ai/v0/tts" \
2
  -H "Content-Type: application/json" \
3
  -H "X-Hume-Api-Key: $HUME_API_KEY" \
4
  -d '{
5
    "utterances": [
6
      {
7
        "text": "Welcome to my application!",
8
        "description": "A friendly and professional voice"
9
      }
10
    ]
11
  }'

Using saved voices

If you have saved voices in your library, reference them by name or ID instead of the model generating the voice. You can still provide descriptions to adjust how the voice performs the text:

1
curl -X POST "https://api.hume.ai/v0/tts" \
2
  -H "Content-Type: application/json" \
3
  -H "X-Hume-Api-Key: $HUME_API_KEY" \
4
  -d '{
5
    "utterances": [
6
      {
7
        "text": "Welcome to my application!",
8
        "voice": {
9
          "name": "MyCustomVoice"
10
        }
11
      }
12
    ]
13
  }'

Advanced options

Speech consistency

For longer content or multiple requests, use the context parameter to maintain consistent speech style. You can provide previous utterances or a generation ID as context:

1
curl -X POST "https://api.hume.ai/v0/tts" \
2
  -H "Content-Type: application/json" \
3
  -H "X-Hume-Api-Key: $HUME_API_KEY" \
4
  -d '{
5
    "utterances": [
6
      {
7
        "text": "The next chapter begins."
8
      }
9
    ],
10
    "context": {
11
      "generation_id": "91f7c154-frz2-4617-8edb-e9n1pg6sce64"
12
    }
13
  }'

Multiple generations

Request up to 5 variations of synthesized speech by setting num_generations. This is useful when you want to explore different interpretations of your voice description:

1
curl -X POST "https://api.hume.ai/v0/tts" \
2
  -H "Content-Type: application/json" \
3
  -H "X-Hume-Api-Key: $HUME_API_KEY" \
4
  -d '{
5
    "utterances": [
6
      {
7
        "text": "Welcome to my application!"
8
      }
9
    ],
10
    "num_generations": 3
11
  }'

Response format

The API returns a JSON response containing:

1
{
2
  "generations": [
3
    {
4
      "audio": "//PExAAspDoWXjDNQgk3HJJJZNbaEPZMmTJk7QIECBA...",
5
      "duration": 1.589667,
6
      "encoding": {
7
        "format": "mp3",
8
        "sample_rate": 24000,
9
      },
10
      "file_size": 26496,
11
      "generation_id": "41f7c154-fbb2-4372-8ecc-e6b7bf6ace01",
12
      "snippets": [
13
        {
14
          "audio": "//PExAAspDoWXjDNQgk3HJJJZNbaEPZMmTJk7QIECBA...",
15
          "id": "37a108c4-5de7-4507-8a54-0521f5cb0383",
16
          "text": "Welcome to my application!" 
17
        }
18
      ]
19
    }
20
  ],
21
  "request_id": "7903e4a7-6642-491a-aa96-c6b359dd1042707439"
22
}

Each response includes:

• Base64-encoded audio in your specified format (MP3, WAV, or PCM).
• A unique generation_id for saving voices or maintaining consistency.
• Audio metadata including duration, file size, and audio encoding.
• A list of segmented utterances (segments) divided into natural sounding units.
• A request_id for tracking and debugging.

Quickstart

Accelerate your project setup with our comprehensive quickstart guides, designed to integrate Octave TTS into your TypeScript or Python applications. Each guide walks you through API integration and demonstrates text-to-speech synthesis, helping you get up and running quickly.

TypeScript

Integrate Octave TTS into web and Node.js applications using our TypeScript SDK.

Python

Use our Python SDK to integrate Octave TTS into your Python applications.

CLI

Get started synthesizing text-to-speech with our command-line tool.

API limits

• Request rate limit: 50 requests per minute
• Maximum text length: 5,000 characters
• Maximum description length: 1,000 characters
• Maximum generations per request: 5
• Supported audio formats: MP3, WAV, PCM