PyTorch vs TensorFlow vs JAX
A side-by-side look at how actively PyTorch, TensorFlow and JAX are developed on GitHub — popularity, contributor base, commit activity and how quickly issues and pull requests move.
What the numbers say
- TensorFlow is the most-starred of the 3, with 201k stars — about 5.5× JAX's 36k.
- Over the 52 weeks to the snapshot, PyTorch recorded the most commits (18k), against 7.8k for JAX.
- PyTorch has closed or merged the most pull requests in its lifetime: 134k.
- Relative to its popularity, TensorFlow carries the lightest open-issue load: 3.2 open issues per 1,000 stars, versus 135.8 for PyTorch.
Head-to-head
Snapshot taken 30 September 2026.
| Metric | PyTorch Slowing | TensorFlow Active | JAX Active |
|---|---|---|---|
| Stars | 104k | 201k | 36k |
| Forks | 31k | 78k | 3.8k |
| Contributors | 7.1k | 5.3k | 1.1k |
| Commits, last 52 weeks | 18k | 14k | 7.8k |
| Open issues | 14k | 647 | 1.8k |
| Open pull requests | 3.5k | 2.6k | 842 |
| Closed / merged PRs | 134k | 80k | 29k |
| Watchers | 1.8k | 7.5k | 333 |
| Last push | 30 Sep 2026 | 30 Sep 2026 | 30 Sep 2026 |
| Latest release | v2.14.0 (Sep 2026) | v2.21.0 (Mar 2026) | jax-v0.11.2 (Sep 2026) |
| Main language | Python | C++ | Python |
| Licence | Not detected by GitHub | Apache-2.0 | Apache-2.0 |
| Created | Aug 2016 | Nov 2015 | Oct 2018 |
Bold marks the highest value in a row. “Active / Slowing / Dormant” is a simple heuristic from how recently code was pushed and open issues per star — a starting point, not a verdict. Stars measure popularity, not quality.
Open these in the live comparer →