Revocalize AI – API Documentation 🎙️

Revocalize AI is a cutting-edge voice platform providing state-of-the-art voice modeling and fingerprinting technology to generate natural-sounding singing voice models. With millions of hours of audio training data, you can now sing like any singer using AI and instantly record & convert your voice into any other voice. ⚡️

Explore the power of Revocalize AI and join our growing community of developers, artists, music producers, audio engineers & enthusiasts.

Getting started • Live Demo • Installation • Pricing • Contact Us

🚀 API Features

The Revocalize AI API offers a variety of features:

• 🌍 Works in any language: Your voice model retains the original accent, tone, and pronunciation, irrespective of the language.
• 🎭 Emotional range: Express true emotions through the voice, from joy and excitement to sadness, and more.
• 🎶 Real-Time Auto Tune: Hits those high notes with ease, thanks to real-time vocal tuning.
• 🔄 Auto-generate variations: Create multiple unique voice variations, each carrying its own unique emotion.
• 🔊 Voice modulation: Adjust pitch, volume, and speed of singing or speech to produce more pleasing results.

🔑 Setup Authentication

To use the Revocalize API, you'll first need to generate an API Key through your Revocalize dashboard at www.revocalize.ai/api-keys. Tokens must be included in the Authorization HTTP header in all requests as a type Bearer.

'Authorization': 'Bearer <revocalize_api_key>'

You can test if everything is working correctly using the ping endpoint and curl:

curl -H "Authentication: Bearer " https://api.revocalize.ai/ping

The response will be {"ping":"pong"} (if it all goes well) 🏓

🛠️ API Endpoints

The API provides multiple endpoints to interact with Revocalize's voice modeling and fingerprinting technology.

POST Convert Audio File 🔄

This endpoint takes a WAV vocal audio file and returns a URL string with the converted vocals audio file.

curl --request POST \
      --url https://api.revocalize.ai/convert \
      --header 'Authorization: Bearer <token>' \
      --data '{
      "audio_file": "<audio_file>",
      "model": "<model>" }'

The response will be a task_id string, which is the ID of the conversion task.

POST Create Model 📦

This endpoint creates a new AI voice model, which is initially created in a pending state. You must train the model before it can be used for synthesis.

curl --request POST \
      --url https://api.revocalize.ai/models/create \
      --header 'Authorization: Bearer <token>' \
      --data '{"training_audio_files": "<training_audio_files>"}'

You'll get a model_id string as a response, which is the ID of the pending model. ⏳

POST Train Model 🏋️‍♀️

Train a custom AI voice model after creating it in a pending state.

curl --request POST \
      --url https://api.revocalize.ai/models/{model_id}/train \
      --header 'Authorization: Bearer <token>' \
      --data '{"epochs": 0}'

The response will include the status of the training task 🚦, the ID of the model being trained, and the current epoch.

Example Request:

curl --location --request POST 'https://api.revocalize.ai/models/123456/train' \
--header 'Content-Type: application/json' \
--header 'Authorization: Bearer <token>' \
--form 'epochs=100'

Response:

{
  "status": "training",
  "model_id": "123456",
  "current_epoch": 50001
}

📪 Contact

If you have any questions, suggestions, or need assistance, feel free to reach out to us:

• Email: [email protected]
• Twitter: @Revocalize
• GitHub: Revocalize AI

We're always happy to help and hear your feedback!

📄 Copyright

Copyright © 2023 Revocalize AI. All rights reserved.

Unauthorized copying, distribution, or use of this documentation or any part of it is strictly prohibited without the express permission of Revocalize AI. Please contact [email protected] for any inquiries or permissions.