Back to Gap Map Model components / Federated learning

TensorFlow Federated

open source / Overall score: 1.5

Google's open framework for machine learning and other computations on decentralized data. It provides a language for expressing federated computations, a Federated Learning API with aggregators, and a single-machine simulation runtime for experiments.

The repository moved from tensorflow/federated to google-parfait/tensorflow-federated; PyPI's project URLs name the new home. The newest PyPI release, 0.87.0, is dated 2024-09-17, while the repository was pushed 2026-09-26.

Openness

5 high confidence

5.0

license: Apache-2.0(OSI
source: public(the whole framework)
core features withheld: no — a research framework with no paid edition

TensorFlow Federated is Apache-2.0 and was built for open research, with no paid tier or held-back component.

https://raw.githubusercontent.com/google-parfait/tensorflow-federated/HEAD/LICENSE recorded 2026-09-27
The LICENSE body: the verbatim Apache License, Version 2.0 text with no added clause.
https://raw.githubusercontent.com/google-parfait/tensorflow-federated/HEAD/README.md recorded 2026-09-27
README: TFF "has been developed to facilitate open research and experimentation with Federated Learning". No paid tier is named.
https://ungh.cc/repos/google-parfait/tensorflow-federated recorded 2026-09-27
Repository record: google-parfait/tensorflow-federated, "An open-source framework for machine learning and other computations on decentralized data", pushed 2026-09-26.
https://ungh.cc/repos/google-parfait/tensorflow-federated/files/main recorded 2026-09-27
The 907-file tree of google-parfait/tensorflow-federated at `main`: one root LICENSE, and no ee/, enterprise/, commercial/, proprietary/ or premium/ path anywhere in it. Path absence only, paired with the statement cited beside it.

Adoption

1 medium confidence

1.0

PyPI downloads of `tensorflow-federated`, the framework itself.

https://pypi.org/pypi/tensorflow-federated/json recorded 2026-09-27
PyPI `tensorflow-federated` 0.87.0, uploaded 2024-09-17; project URLs point at github.com/google-parfait/tensorflow-federated.
https://pypistats.org/api/packages/tensorflow-federated/recent recorded 2026-09-27
last_month = 2,298 downloads for the package `tensorflow-federated`.

Capability

2 high confidence

2.0

TFF expresses federated computations well, but what it ships is a simulation runtime, and its own deployment guide calls simulation the primary focus and is unfinished. That puts it level with FedJAX.

https://raw.githubusercontent.com/google-parfait/tensorflow-federated/HEAD/docs/deployment.md recorded 2026-09-27
Deployment: "Whereas the primary focus is on simulations, the interfaces and tools we provide are more general"; "Note: This document is still under construction."
https://raw.githubusercontent.com/google-parfait/tensorflow-federated/HEAD/README.md recorded 2026-09-27
README: a "single-machine simulation runtime for experiments".

Verified 2026-09-27