Updated 20 Oct 2026: Add October 8 Effect 4.0.2 correction: accidental stable-API promise narrowed via explicit unstable/experimental annotations; no current code/signature changes. Clarify distinction from three-year LTS.

Key details

  1. Effect 4.0 released September 30, 2026.
  2. The core effect package has zero runtime dependencies.
  3. Vendor reports minimal bundles shrinking from 35.6 KB to 7.1 KB.
  4. Vendor reports throughput rising 0.71M to 4.57M tasks/s and 86% lower memory for 50,000 fibers.
  5. 4.x has a minimum support floor through September 2029.
  6. Some newer modules remain unstable or experimental.
  7. Effect 4.0.2 (October 8) retroactively marks some 4.0.0 APIs unstable/experimental; current signatures and behavior were not changed.
  8. The maintainers explicitly characterize the corrected stability promise as technically breaking and added CI stability-leak checks.

What builders should take away

  1. Use the official migration guide and run application-specific type/build tests.
  2. Rebenchmark memory and concurrency under real workloads before planning capacity reductions.
  3. Review consolidated package imports and dependency lockfiles.
  4. Record the LTS support window in your maintenance policy.
  5. Inspect the API reference and language-service warnings for unstable/experimental exports after updating to 4.0.2 or later; do not equate major-version LTS with a stable guarantee for every module.

What changed

Effect 4.0 shipped September 30, 2026 as a major rebuild. It folds formerly separate packages into a shared-version ecosystem, removes runtime dependencies from the core effect package, and publishes vendor benchmarks comparing a 35.6 KB to 7.1 KB minimal program, 0.71M to 4.57M tasks/s and 157.5 MB to 21.8 MB heap for 50,000 fibers. The project also introduces a formal LTS policy: Effect 4 receives bug and security fixes until at least September 2029, with extensions tied to the timing of a future 5.0 release. On October 8, the maintainers disclosed that 4.0.0 had shipped several intended-unstable modules without the required @stability tags. Under their original default-stable policy, this accidentally promised stability. Effect 4.0.2 added the missing unstable/experimental annotations and automated CI checks for stability leakage. The team calls the reduced compatibility promise technically breaking, but says it did not change code behavior, API signatures or types.

Why it matters

Effect underpins typed errors, resource management, structured concurrency and workflows in TypeScript services. Changes to its runtime, package graph and support lifecycle affect build size, memory budgets, dependency security and long-term maintenance planning. The support policy is an operational commitment; the performance claims are benchmarks under defined vendor conditions, not universal results. The October 8 correction is especially important for teams selecting APIs on the basis of long-term compatibility: a three-year supported major line does not mean every new module in it has a stable API contract. Check per-export annotations before depending on recently introduced features.

A smaller zero-dependency core

Effect 4 consolidates packages into the effect distribution and removes third-party runtime dependencies from its core, reducing dependency-chain and bundle overhead.

Vendor benchmarks show large gains under defined tests

The team reports 5x smaller minimum bundles, 6.4x concurrent task throughput and 86% lower memory per fiber. Benchmarks use identical source for bundles and nine fresh-process runs for runtime medians.

LTS becomes an explicit contract

The 4.x line receives bug/security fixes until at least September 2029; depending on the next major release date, support can extend beyond that floor.

Migration and unstable modules still need review

A migration guide exists, but some new modules remain marked unstable or experimental and may change in smaller releases.

4.0.2 corrected accidentally promised API stability

In its October 8 post, the Effect team acknowledged that some modules meant to remain unstable had shipped in 4.0.0 without explicit @stability tags. The stated policy had treated untagged exports as stable, making the correction a narrowing of a public compatibility commitment. Effect 4.0.2 adds annotations rather than changing runtime behavior, types or signatures. The project now requires explicit stability declarations and runs CI diagnostics to catch stable APIs leaking unstable types. The Effect language service warns when a project uses unstable or experimental APIs, with granular opt-ins available through allowedUnstableApis and allowedExperimentalApis.

What to watch next

  • Independent production benchmark results and regressions.
  • Effect 4 module stability promotions.
  • Timing and support-policy details for Effect 5.
  • Whether further API stability annotations or breaking compatibility clarifications follow the 4.0.2 correction.

Still unclear

  • Performance figures are vendor measurements, not independent cross-application guarantees.
  • Some modules remain experimental and may evolve.
  • Support extensions depend on the future 5.0 release date.
  • The primary October 8 post describes categories of newly tagged modules but does not quantify how many production applications depend on them.

Sources

Direct reading behind this dossier.

3 sources
Effect 4.0
Effect primary

Primary architecture, benchmark methods and LTS commitments.

Effect 4.0: The First Week
Effect primary

Maintainer admission of accidentally stable API contracts, 4.0.2 annotations, no runtime/signature changes, and CI/language-service protections.

Discussion

Discussion is reader-contributed. Comments are not part of the BTN dossier or its editorial evidence.

0 visible comments

Join the discussion

Keep comments useful and relevant. Reader contributions may be moderated and are not BTN editorial evidence.

Sign in to comment