PaddlePaddle

license

Apache-2.0(OSI)

source

public(PaddlePaddle/Paddle)

core-gated

ungated(no enterprise or ee path in the repository root and no paid tier described in the README)

Apache-2.0, an OSI licence, with the full source public and the shipped code in this repository. The LICENSE body was read in full. A copyright line for the PaddlePaddle Authors is prepended above the licence text, which is why an automated reader can report the file as non-standard; the remainder is byte-identical to the canonical Apache-2.0 template and nothing follows END OF TERMS AND CONDITIONS but the standard appendix. No enterprise or ee path in the repository root and no paid tier described in the readme.

• https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/LICENSE recorded 2026-08-31

  LICENSE body, read in full: a prepended 'Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserved' line above the stock Apache License 2.0 text. The remainder is byte-identical to the canonical template through END OF TERMS AND CONDITIONS, followed only by the standard appendix. No appended condition.

• https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/README.md recorded 2026-08-31

  README, read for the distribution and gating question: it documents installing the framework from this repository or from the paddlepaddle package and names no enterprise, paid or licence-gated build.

Banded on the authoritative PyPI route: 1,785,666 downloads in the last 30 days for the paddlepaddle package, which places it at level 4. The repository carries 24,064 stars, but stars rank below every real usage signal and are recorded here only as corroboration.

• https://pypistats.org/api/packages/paddlepaddle/recent recorded 2026-08-31

  last_month 1785666, last_week 240058, last_day 18483 for the paddlepaddle package.

• https://api.github.com/repos/PaddlePaddle/Paddle recorded 2026-08-31

  Repo metadata for PaddlePaddle/Paddle: stargazers_count 24064, archived false, pushed_at 2026-08-31.

Scored 5 on framework scope. Paddle is a foundational framework in the same sense as pytorch and tensorflow, which both sit at 5 here: it owns its own graph engine, distributed execution and deployment path rather than building on somebody else's. Its automatic-parallelism and integrated training-inference design are the distinguishing claims.

• https://raw.githubusercontent.com/PaddlePaddle/Paddle/develop/README.md recorded 2026-08-31

  The 'PaddlePaddle New Generation Framework 3.2' section documents unified dynamic/static graphs with automatic parallelism from single-card tensor partitioning annotations, integrated training and inference for large models within one framework, and high-order differentiation for scientific computing.