After years of work, months of beta releases, production use across the community, and an in-depth audit of the entire codebase, we think Effect is ready for 4.0.
This is a milestone we have been working toward for a long time, and we are incredibly excited and proud about what we are about to ship. It will be our best release yet.
We just tagged our release candidate.
Now it’s your turn to prove us right, or give us your veto before we ship Effect 4.0.
npm install effect@rcpnpm add effect@rcyarn add effect@rcbun add effect@rcWhat release candidate means
This release candidate is a statement of confidence. It’s not a claim that there is no more work to do.
We are confident that the sweeping changes are behind us, and we have no more broad breaking changes planned. At this stage, any breaking change that we might make must address something we strongly believe would be wrong to ship otherwise.
In other words: our interfaces are now presumed final.
Our focus between now and stable is validation: fixing regressions, compatibility issues, performance problems, bugs, and rough edges that emerge through our and your testing. If a narrowly scoped breaking change proves necessary, we will communicate it clearly and provide a straightforward migration path.
Why we believe it is ready
Many projects in the Effect community have already migrated during our beta and have been running it in production for months. Their experience exposed cases our examples and tests could not.
Additionally, we spent the past two months auditing the entire codebase. We fixed all the subtle issues and edge cases we found, expanded the tests around them, and kept auditing until our new findings became increasingly esoteric or died away. Now we want a much larger set of real applications to take the next swing at it.
Now is the best time to migrate
If you have an existing Effect project, now is the best time to migrate.
Ensuring a smooth transition is the team’s highest priority. The whole team is actively working on migration support, developer experience, agent experience, and further hardening. We are monitoring GitHub issues closely and will respond quickly to bugs, migration blockers, and papercuts.
The migration instructions cover package changes, interface renames, core behavioral changes, and other important parts relevant to the transition. They are intended to give your coding agent enough context to perform a near-fully automated migration.
Hand these instructions to your coding agent, review the resulting changes, and run your test suite. If the agent gets something wrong, cannot complete part of the migration, or needs information that the guide does not provide, please tell us. Feedback about the migration instructions and the agent-assisted migration experience is especially valuable right now.
For additional migration guidance or questions, join us on Discord. The maintainers are active there, and the wider community is generous with its time and experience.
We will also theme our weekly Effect Office Hours around this phase. Join us live to ask maintainers and community guests about the migration and any problems you encounter while adopting Effect.
Your call to action
Our test suite cannot reproduce every application, integration, deployment environment, or production workload. But yours can.
We want everyone in the Effect community to start migrating and validating against real applications. Run your builds and test suites, observe, and tell us what breaks. Look out for regressions, bugs, migration gaps, papercuts or performance issues.
If you find something that should stop us from shipping, we want to know now.
Remember that you are the final gate between this and stable. It’s your turn now. Don’t hold back. We need you to ruthlessly evaluate it from every possible angle.
We think we are ready, but we need you to verify.
What’s next
The quality of the library itself is obviously only one part of the desired experience.
Documentation
Effect has a huge scope, and it deserves documentation with matching quality and coverage. Over the coming weeks, we will be deeply invested in a thorough rewrite and in building the documentation experience that Effect deserves. In other words: we have set a very high bar for ourselves.
Tooling
The same applies for our tooling. The developer and agent experience is not yet where we want it to be. Over the coming weeks, we will invest deeply in language tooling and tools that help developers and coding agents understand, write Effect code. We are aiming for a tooling experience that matches the scope and ambition of Effect itself, and we will share more when it is ready.
Last but not least
We are targeting the stable release for Q3/Q4 2026.
Thank you for your attention to this matter.