ActiveState Platform vs Hatch

ActiveState Platform hosts custom runtimes assembled from source. Projects specify runtime components, preserve project history and provide deployment instructions; their artifacts may include packages, dependencies, patches and bundles, but not contributors’ first-party code. Create projects in the web interface or with the State Tool CLI, then deploy and activate them locally. Projects can remain personal, be shared with an organization or be public. The platform supports Windows, macOS and Linux deployments, and its GraphQL API can build open-source software for different operating systems and languages without requiring users to manage build infrastructure. It provides immutable SBOMs in SPDX formats and describes build safeguards including reproducible builds, checksum checks and isolated containers. A free organization plan costs nothing and is limited to public projects. Curated Catalog annual pricing is $25,000 for fewer than 100 employees, $50,000 for 100–999 employees and $150,000 for 1,000 or more; those prices do not apply to legacy Platform subscriptions.

Hatch is a free, MIT-licensed Python project manager for creating projects and managing development environments on Linux, macOS, and Windows. It can start projects from templates or initialize existing ones, including migrating setuptools configuration from setup.py. Environment management supports custom scripts and UV, with manual Python installations or automatic installation as part of an environment. Hatch also supports reproducible builds, test execution, and static analysis backed by Ruff. For publishing, it can upload packages to PyPI and other package indices; lockfiles, dependency groups, and private registries are listed capabilities. Its plugin system uses pluggy, with documented extension types for environments, builders, publishers, and other project tasks. Visual Studio Code can discover Hatch environments through the vscode-python extension, and an official GitHub Actions install action is documented. Installation options include standalone installers or binaries, pip, pipx, Homebrew, Conda, and other methods. The documentation provides a self-report command that prepares a GitHub issue for unexpected behavior.