---
type: "Framework"
title: "Flower"
description: "Framework-agnostic federated AI: simulation-to-deployment, secure agg, DP, audit, OIDC."
resource: "https://flower.ai/"
tags: [fl, framework]
generated: { by: human:crpage, at: 2026-07-09T09:44:00Z }
vendor: "Flower Labs"
status: stable
sources: [{ id: primary, resource: "https://flower.ai/" }]
---

An open-source federated AI framework with a clean split between **simulation** and **deployment** runtimes (SuperLink, SuperExec, ServerApp, SuperNodes). Framework-agnostic across PyTorch, TensorFlow, Hugging Face, scikit-learn, JAX, XGBoost. Public docs show secure aggregation, differential privacy, TLS, SuperNode authentication, enterprise audit logging and OIDC/OpenFGA authorisation. A strong **"missing middle"** bridge between research and operations — but not the only answer.

Relates to: [The federation layer](../stack/federation-layer.md), [Flower Labs](../organisations/flower-labs.md).