Voxel51 API Python Client Library

This package defines a Python client library for accessing the Voxel51 Vision Services API.


To install the library, first clone it:

git clone https://github.com/voxel51/api-python
cd api-python

and then install the package:

pip install .

Sign-up and Authentication

To use the API, you must first create an account at https://console.voxel51.com and download an API token. Keep this token private. It is your access key to the API.

Each API request you make must be authenticated by your token. To activate your token, set the VOXEL51_API_TOKEN environment variable in your shell to point to your API token file:

export VOXEL51_API_TOKEN="/path/to/your/api-token.json"

Alternatively, you can permanently activate a token with:

from voxel51.auth import activate_token


In the latter case, your token is copied to ~/.voxel51/ and will be automatically used in all future sessions. A token can be deactivated via the voxel51.auth.deactivate_token() method.

After you have activated an API token, you have full access to the API.

Example Usage

The following examples describe some actions you can take using the API.

To initialize an API session, issue the following commands:

from voxel51.api import API

api = API()


List available algorithms:

algos = api.list_algorithms();

Download algorithm documentation:

doc = api.get_algorithm_doc(algo_id);


  • Upload data to the cloud:
data_metadata = api.upload_data("/path/to/video.mp4")
  • List uploaded data:
data_list = api.list_data()


  • Create a job request:
job_request = JobRequest(algo_id)
job_request.set_input("<input>", data_id=data_id)
job_request.set_parameter("<param1>", val1)
job_request.set_parameter("<param2>", val2)
  • Upload the job request:
job_metadata = api.upload_job_request(job_request, "test-job")
  • List the jobs you have created:
job_list = api.list_jobs()
  • Start a job:
  • Get the status of a job:
job_status = api.get_job_status(job_id)
  • Download the output of a completed job:
api.download_job_output(job_id, output_path)

Generating Documentation

This project uses Sphinx-Napoleon to generate its documentation from source. To install the necessary dependencies to generate the documentation, run:

pip install --upgrade sphinx
pip install --upgrade sphinx_rtd_theme
pip install --upgrade sphinxcontrib-napoleon
pip install --upgrade m2r

To generate the documentation, run:

bash generate_docs.bash

To view the documentation, open the build/html/index.html file in your browser.