What’s New

Every Anaconda user has hit the same wall. They run conda install, and instead of a package, they get a PackagesNotFoundError. The package is popular on PyPI. But it isn’t in Anaconda’s main channel and now, that means they have to download it from PyPI instead.

For individual developers, this is daily friction. For organizations, it’s a supply chain risk that compounds silently: every package sourced from outside your governed Anaconda channels is a package that bypasses the vulnerability scanning, license validation, and policy enforcement your team put in place. The governance infrastructure exists, but the package gap keeps routing people around it.

That is the gap main-x closes.

Let’s Go Deeper

main-x is a new authenticated Anaconda channel launching today in Open Beta. It adds approximately 6,000 new packages that are ready to work in your existing setup, all built from source by Anaconda with the same checks that users of our main channel expect: license validation, typosquatting protection, and automated testing, including testing to verify compatibility with main packages.

main-x is designed to work alongside main. You log in, add one channel, and thousands more packages become available from the source you already trust. Unlike main, main-x is an authenticated channel hosted at repo.anaconda.cloud. Authentication lets us understand which packages our users need so we can prioritize what to build next. It also ties access to a real Anaconda account, so we can support you if something goes wrong. All you need is a free account; sign up and you’re ready to go.

Why This Matters

Most of the time, when a developer hits a PackagesNotFoundError, they don’t wait for Anaconda to build the package. They pull it from conda-forge, pip-install it directly, or find it on a random GitHub repo, bypassing the vulnerability scanning, license validation, and policy controls their organization built.

The package coverage gap between Anaconda’s main channel and other sources is one of the most common sources of friction we hear from our customers. main-x addresses this directly. By adding 6,000+ Anaconda-built packages to the channels your teams already use, we reduce the number of times a developer has to make that choice between waiting or falling back to conda-forge and PyPI. The packages come from the same infrastructure as main, with the same build standards, so when a developer installs from main-x their organization’s supply chain controls stay intact.

The gap closes by giving developers more packages, secure by default, with the same trusted build standards they already rely on.

Give It a Shot

Make sure you have anaconda-auth version 0.14.4 (or later) installed in your base environment. For full setup instructions (including those for Business tier and Anaconda Platform customers), see our documentation.

main-x is free to access with any Anaconda account. Create a free account, authenticate with anaconda login, and add the channel:

anaconda login
conda config --append channels https://repo.anaconda.cloud/repo/main-x

What To Expect in Beta

Pure Python packages only. The initial ~6,000 packages are pure Python. Compiled packages are out of scope for Beta. Coverage grows over time, driven by user demand.

No CVE information or SBOMs yet. Build standards are the same as main, but vulnerability and SBOM metadata aren’t available in Beta. If your organization requires vulnerability data or SBOMs before approving packages, wait for general availability (GA), coming in Q3.

Not for production use. Beta is for development and evaluation. We’ll be explicit when production use is supported.

No service account support. CI/CD pipelines requiring non-interactive authentication won’t work in Beta. That’s planned for GA.

Tell Us What You Need

Submit missing package requests under Packages & Installers. Include the package name, version, and your platform (e.g., Ubuntu 22.04, macOS Sonoma on Apple Silicon, Windows 11).

For bugs or technical issues, please open a request through the Anaconda Support Center with:

  • A description of the issue or question
  • The command(s) you ran and the full error message (if applicable)
  • Your operating system and platform (e.g., Windows 11, macOS Sonoma on Apple Silicon, Ubuntu 22.04)
  • Your conda version (conda --version)