# NVIDIA FLARE

> NVIDIA FLARE is a domain-agnostic, open-source, extensible SDK for federated learning and other federated-computing applications, from local development through production deployment.


2025 NVIDIA

## Pages in this subsection

- [ModelController API](model_controller.md): The FLARE `ModelController` API provides an easy way for users to write and customize FLModel-based ...
- [Controllers and Controller API](controllers.md): The `controller API` makes it possible to create any client coordination logic
- [Scatter and Gather Workflow](scatter_and_gather_workflow.md): The federated scatter and gather workflow is an included reference implementation of the default wor...
- [Cross Site Model Evaluation / Federated Evaluation](cross_site_model_evaluation.md): The `cross site model evaluation workflow`
- [Cyclic Workflow](cyclic_workflow.md): The cyclic workflow was added in NVIDIA FLARE 2.0 and implemented with the Controller API. It allows...
- [Initialize Global Weights Workflow for Client-Side Global Model Initialization](initialize_global_weights.md): The SAG controller requires the global model weights to be initialized before the training is starte...
- [Client Controlled Workflows](client_controlled_workflows.md): Server-based controlling usually assumes that the server is trusted by all clients since results sub...

## Optional

- [Top-level llms.txt](../../llms.txt): Complete documentation index.