Workflow Blocks

Explore the ecosystem of Blocks you can use in a Workflow.

Workflows are made up of Blocks. You can connect multiple Blocks together to build computer vision applications.

In this section of the Roboflow documentation, we will cover:

  1. What each Block does

  2. How to use each Block

  3. When a Block may be useful

  4. The response returned by a Block

This section is intended as reference as you build a Workflow. We encourage you to come here whenever you have a question about how to use a particular Block in the Workflows editor.

Several categories of Blocks are available:

Category
Descriptoin

Models

Run a fine-tuned or foundation model.

Visualizations

Visualize the output of a model.

Logic and Branching

Control the flow of your Workflow.

Data Storage

Save data in a Roboflow dataset or an external database.

Notifications

Send an alert, such as an SMS message or email.

Video Processing

Analyze the contents of videos with tracking algorithms, time in zone analysis, and line crossing checks.

Transformations

Manipulate image and prediction data.

Classical Computer Vision

Run classical computer vision tasks, such as edge deteciton, template matching, and size measurement.

Enterprise

Connect Roboflow to your Enterprise systems, such as using MQTT or Modbus TCP.

Advanced

Advanced blocks for specific use cases, like PASS/FAIL analysis, data caching, and embedding similarity.

Custom

Make and run custom blocks with Python

Last updated

Was this helpful?