AI-Hypercomputer/JetStream

Model Serving

JetStream is a throughput and memory optimized engine for LLM inference on XLA devices, starting with TPUs (and GPUs in future -- PRs welcome).

GitHub →

4.8

GitHub Metrics

Stars

455

Forks

Open Issues

Watchers

Contributors

Weekly Commits

Language

Python

License

Apache-2.0

Last Commit

Jan 5, 2026

Created

Mar 1, 2024

Latest Release

v0.3

Release Date

Dec 18, 2024

Synced: Aug 28, 2026

Quality Scores

Documentation Qualityw: 20%

4.8

No dedicated docs site. Description: 143 chars. Stars signal: 455. Contributors: 38. Score: 4.8/10

Community Healthw: 20%

5.4

Stars: 455. Contributors: 38. Watchers: 22. Forks: 67. Issue ratio: 5.7%. Score: 5.4/10

Maintenance Velocityw: 15%

2.4

Last commit: 234d ago. Weekly commits: 0. Latest release: v0.3. Maturity bonus: 2.5y old. Score: 2.4/10

API Design & DXw: 20%

5.8

Stars/issues ratio: 18. Dynamic language: Python. No dedicated API docs. Permissive license: Apache-2.0. Popularity signal: 455 stars. Score: 5.8/10

Production Readinessw: 15%

4.0

Battle-tested: 455 stars. Peer review: 38 contributors. Versioned: v0.3. Licensed: Apache-2.0. Age: 2.5 years. Maintenance: last commit 234d ago. Score: 4/10

Ecosystem Integrationw: 10%

6.2

Fork interest: 67. Major ecosystem: Python. Integration-friendly: Apache-2.0. Adoption: 455 stars. Score: 6.2/10