resemble.ai API

resemble.ai is a state-of-the-art natural voice cloning and synthesis provider. Best of all, the platform is accessible by using our public API! Sign up here to get an API token!

This repository hosts a Python library for convenient usage of the Resemble API.

Quick start

from resemble import Resemble

Resemble.api_key('your_api_key')

project = Resemble.v2.projects.get('project_uuid')
voice = Resemble.v2.voices.get('voice_uuid')

clip = Resemble.v2.clips.create_sync('project_uuid', 'voice_uuid', 'This is a test')

Streaming

The Streaming API is currently in beta and is not available to all users. Please reach out to team@resemble.ai to inquire more.

Streaming example:

from resemble import Resemble

Resemble.api_key('your_api_key')
Resemble.syn_server_url('your_resemble_synthesis_server_url') # Extra configuration required for streaming

for chunk in Resemble.v2.clips.stream('project_uuid', 'vouce_uuid', 'This is a test'):
    # do something with the `chunk` of audio data, such as feeding it into a streaming audio player
    pass

Deepfake Detection

Detect AI-generated or manipulated content in audio, video, and image files.

from resemble import Resemble

Resemble.api_key('your_api_key')

# Asynchronous detection (default)
detection = Resemble.v2.deepfake_detection.detect(
    url='https://example.com/suspicious_audio.wav',
    callback_url='https://your-site.com/webhook',  # Optional: receive results via webhook
    visualize=True  # Optional: generate detection visualizations
)

print(f"Detection started with UUID: {detection['item']['uuid']}")

# Check detection results
results = Resemble.v2.deepfake_detection.get(detection['item']['uuid'])
print(f"Detection status: {results['item']['metrics']}")

# Synchronous detection (wait for results)
sync_results = Resemble.v2.deepfake_detection.detect(
    url='https://example.com/video.mp4',
    synchronous=True,
    frame_length=2,  # Analysis frame length in seconds
    pipeline='general->facial',  # Detection pipeline for images/video
    max_video_fps=30  # Limit video processing FPS
)

Supported formats:

• Audio: wav, mp3, ogg, m4a, flac
• Video: mp4, mov, avi, wmv
• Image: jpg, jpeg, png, gif, webp

Development

The library files are located in resemble/

Testing

Note that currently the test suite performs real actions against the Resemble AI platform (e.g., creating projects, retrieving clips, updating voices, et cetera).

# For tests that do NOT require a voice
$ TEST_API_KEY=<...> TEST_BASE_URL=<...> python3 -m unittest

# For tests that do require a voice
$ TEST_API_KEY=<...> TEST_BASE_URL=<...> TEST_VOICE_UUID=<...> python3 -m unittest

Publishing new versions

You need the build and twine pip packages. (pip3 install build twine)

1. git status: Make sure your working directory has no pending changes.
2. Update the version attribute in setup.py.
3. git commit: Commit this version change.
4. Build the package: make pkg.build
5. (optional) Publish to the test index:

make pkg.publish.test

6. Publish to the index

make pkg.publish