#4141st of 75 in ML frameworks and libraries
metaflow
Python framework for building, scaling and deploying ML and AI workflows
Live demo ↗
(opens in a new tab)Documentation ↗
(opens in a new tab)Website ↗
(opens in a new tab)Repository on GitHub ↗
(opens in a new tab)
- Stars
- 10.3k
- License
- Apache-2.0
- Last commit
- Oct 2026
- Last release
- Sep 2026
- Language
- Python
9,933 on 31 Mar 202610,308 on 20 Oct 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
Metaflow is a Python library for writing ML and AI pipelines as flows of steps, which run locally or in a notebook and then scale out to cloud compute on CPUs and GPUs. It tracks experiments, versions artifacts, manages dependencies, and deploys flows to production orchestrators with event-based triggering. Originally built at Netflix, now supported by Outerbounds.
Who it is for: ML and data engineers moving Python pipelines from notebooks to production
Strengths
- Same flow code runs on a laptop, in a notebook, on cloud compute and in production
- Built-in experiment tracking, artifact versioning and a client API for inspecting past runs
- Supports foreach parallelism and gang-scheduled distributed jobs on CPUs and GPUs
- Apache-2.0, installable from PyPI or conda-forge
Weaknesses
- Cloud scaling and production orchestration need separate infrastructure setup beyond pip install
- README does not list which cloud providers or orchestrators are supported
- Not LLM-specific; README describes no model serving or agent features
- No Docker or compose setup in the repository
What it needs
- GPU optional
- Needs Python, Cloud compute cluster (for scaling), Production workflow orchestrator (for deployment)
Also in ML frameworks and libraries
See all 75| Rank | Project | Score |
|---|---|---|
| 1 | transformersPython library defining open models for text, vision, audio and multimodal tasks | 94 out of 100 |
| 2 | pytorchPython tensor library with GPU support and tape-based autograd for deep learning | 94 out of 100 |
| 3 | ultralyticsPython library and CLI for training and running YOLO vision models | 92 out of 100 |
| #4 | pytorch-image-modelsPyTorch library of image models, pretrained weights and reference training scripts | 91 out of 100 |
| #5 | supervisionPython toolkit for loading, annotating and counting computer vision detections | 90 out of 100 |