About CML
CML, or Continuous Machine Learning, is an open-source command-line tool for applying CI/CD practices to machine learning work, with a focus on MLOps. It is aimed at data science teams who want to automate workflows such as provisioning machines, training and evaluating models, and comparing experiments across a project's history, using the Git-based tools they already rely on.
It can train and evaluate models automatically on every pull request and generate a visual report of results and metrics directly in that pull request, following a 'GitFlow for data science' approach where GitLab, GitHub or Bitbucket track who changed models or data and when, often paired with DVC to version data and models outside of Git itself. CML needs no additional backend services, databases, or complex setup, working on top of existing GitHub Actions, GitLab CI or Bitbucket Pipelines infrastructure with either self-hosted or cloud runners such as AWS EC2 or Azure.
CML is written in JavaScript and released under the Apache-2.0 license. The project provides a Discord community and a YouTube tutorial series for hands-on MLOps guidance, alongside documentation covering setup for each supported Git platform.
Key features
- Automated model training and evaluation in CI
- Auto-generated pull request reports with metrics
- Works with GitHub Actions, GitLab CI, Bitbucket Pipelines
- Integrates with DVC for data and model versioning
- No extra databases or backend services needed
- Self-hosted or cloud runner support
Good fit for
- CI/CD pipelines for machine learning models
- Automated experiment comparison on pull requests
CML: questions and answers
- What is CML used for?
- CML (Continuous Machine Learning) is a free, Apache-licensed CLI tool that brings CI/CD practices to machine learning, automating training, evaluation and reporting. It is a good fit for CI/CD pipelines for machine learning models and automated experiment comparison on pull requests.
- Is CML open source?
- Yes. CML is open source under the Apache-2.0 licence. Its source code is on GitHub at iterative/cml and is written mainly in JavaScript.
- Is CML free?
- Yes. CML is open source, so the software itself is free to use.
- What is CML an alternative to?
- CML is an open-source alternative to Valohai. Other open-source alternatives to Valohai include labml, MLflow and Flyte.
- Is CML actively maintained?
- Yes. The most recent commit to CML was on 21 October 2026, and the latest release is v0.20.6, published on 24 October 2024. The project has 4.2k stars on GitHub, with 1 added in the last 30 days.
Open-source alternatives to CML
See all
labml
AI Infrastructure
🔎 Monitor deep learning model training and hardware usage from your mobile phone 📱
MITvs Weights & Biases★ 2.3k
MLflow
AI Infrastructure
The open source AI engineering platform for agents, LLMs, and ML models. MLflow enables te
Apache-2.0vs Weights & Biases★ 29k
Flyte
AI Tools
Type-safe, distributed orchestration of agents, ML pipelines, and real-time inference on y
Apache-2.0vs Amazon SageMaker★ 7.8k
Kubeflow
AI Tools
The Cloud Native AI Platform
Apache-2.0vs Together AI★ 16k
Polyaxon
Workflow Automation
AI Infra / AI Orchestration / AI Control Plane
Apache-2.0vs Amazon SageMaker★ 3.7k
Pachyderm
Data Pipelines & ETL
Data-Centric Pipelines and Data Versioning
Apache-2.0vs Databricks★ 6.3k
SaaS alternatives to CML
See all
Valohai
AI Infrastructure
MLOps platform for automating machine learning pipelines and experiment tracking
SaaS
AppVeyor
CI/CD & DevOps
CI/CD service for Windows, Linux and macOS projects
SaaS
AWS CodeBuild
CI/CD & DevOps
Managed AWS build service that compiles source code and runs tests
SaaS
Azure Pipelines
CI/CD & DevOps
CI/CD service in Azure DevOps for building and deploying to any platform
SaaS
Blacksmith
CI/CD & DevOps
Faster GitHub Actions runners billed per minute, with build caching
SaaS
CircleCI
CI/CD & DevOps
Hosted continuous integration and delivery platform for build and test pipelines
SaaS

