Verified project record

roboflow/supervision

Supervision provides reusable Python building blocks for computer vision workflows, including model-agnostic connectors for detection and segmentation results and customizable visualization annotators. It is intended for developers who need to reduce repetitive code for data loading, detection processing, and visualization across image and video inputs.

Project overview

Project type: Model Development · Image & Vision · Video
Deployment: Refer to project documentation
License: MIT

Best for

Developers and AI engineers who want to process image or video streams locally without a paid service and need to visualize or count detections from supported models.

Key capabilities

Provides connectors to plug in and use results from various classification, detection, or segmentation models.
Offers a range of customizable annotators to compose visualizations for bounding boxes and other detections.

Limitations and risks

Requires a Python environment of version 3.10 or higher.
GPU and minimum hardware requirements are not documented.

Getting started

Installs cleanly with a standard pip command without requiring complex system dependencies.

Evidence and sources

GitHub project description: We write your reusable computer vision tools. 💜
README: **We are your essential toolkit for computer vision.** From data loading to real-time zone counting, we provide the building blocks so you can focus on building applications aroun…
README: Pip install the supervision package in a [**Python>=3.10**](https://www.python.org/) environment. ```bash pip install supervision ```
README: Supervision was designed to be model agnostic. Just plug in any classification, detection, or segmentation model. For your convenience, we have created [connectors](https://superv…
README: Running with [Inference](https://github.com/roboflow/inference) requires a [Roboflow API KEY](https://docs.roboflow.com/api-reference/authentication#retrieve-an-api-key).

Reviewed at: 2026-07-17T07:09:57Z