What's new in beta.35

May 8, 2026

v2.0.0-beta.35 is out. The headline addition is an Effect-native, Stream-shaped consumer API for Cloudflare Queues — write a queue handler the same way you’d write any other Effect pipeline. Plus R2 bucket custom domains, non-string Worker env bindings, Neon logical replication, and a handful of CLI polish. Community contributors get inline shoutouts; full credits in the Contributors section.

`Cloudflare.messages(queue).subscribe(...)` — Effect-native queue consumer

Process Cloudflare Queue batches as a typed Effect Stream, with ack/retry exposed per message. The handler returns when the batch is done; failed messages get retried up to maxRetries. Same Worker, same Effect runtime, same bindings — the queue consumer is just another yield* in the init phase.

A real producer-and-consumer pair: a Worker accepts POSTs to /queue/send, the consumer writes each message to R2 as JSON so the result is observable. Bindings already on the Worker (bucket, queue) are in scope inside the stream handler — the consumer is a peer to the Worker’s init phase, not a separate file.

import * as Cloudflare from "alchemy/Cloudflare";
import * as Effect from "effect/Effect";
import * as Stream from "effect/Stream";
import { Bucket } from "./Bucket.ts";
import { Queue } from "./Queue.ts";

interface QueueMessageBody {
  id: string;
  text: string;
  sentAt: number;
}

export default class Api extends Cloudflare.Worker<Api>()(
  "Api",
  { main: import.meta.filename },
  Effect.gen(function* () {
    const bucket = yield* Cloudflare.R2Bucket.bind(Bucket);
    const queueResource = yield* Queue;
    const queue = yield* Cloudflare.QueueBinding.bind(queueResource);

    // Consumer side. Each batch arrives as a Stream of QueueMessage<Body>;
    // success ack()s every message, failure retry()s. Crash mid-batch and
    // unprocessed messages are redelivered.
    yield* Cloudflare.messages<QueueMessageBody>(queueResource, {
      batchSize: 25,
      maxRetries: 3,
    }).subscribe((stream) =>
      Stream.runForEach(stream, (msg) =>
        bucket
          .put(`/queue/${msg.body.id}`, JSON.stringify(msg.body), {
            httpMetadata: { contentType: "application/json" },
          })
          .pipe(Effect.asVoid),
      ),
    );

    return {
      // Producer side: POST /queue/send adds a message.
      fetch: Effect.gen(function* () {
        const request = yield* HttpServerRequest;
        if (request.url === "/queue/send" && request.method === "POST") {
          const text = yield* request.text;
          const msg: QueueMessageBody = {
            id: crypto.randomUUID(),
            text,
            sentAt: Date.now(),
          };
          yield* queue.send(msg);
          return yield* HttpServerResponse.json({ sent: msg }, { status: 202 });
        }
        return HttpServerResponse.text("ok");
      }),
    };
  }).pipe(
    Effect.provide(
      Layer.mergeAll(
        Cloudflare.R2BucketBindingLive,
        Cloudflare.QueueBindingLive,
        Cloudflare.QueueEventSourceLive, // required for `messages(...)` to dispatch
      ),
    ),
  ),
) {}

The full round-trip lives in examples/cloudflare-worker/src/Api.ts. Tutorial › Queue consumer

R2 bucket custom domains

Attach custom hostnames to an R2 bucket directly from the resource declaration. Add or remove entries by editing the array — Alchemy reconciles the live attachments on the next deploy.

const Assets = yield* Cloudflare.R2Bucket("Assets", {
  customDomains: [
    "assets.example.com",
    "cdn.example.com",
  ],
});

Each domain points at the bucket’s public origin; pair with a Cloudflare Zone if you want the DNS record provisioned in the same stack.

— Thanks to Alex (#241) for the contribution.

Non-string env bindings on Workers

Worker bindings now accept non-string values directly — numbers, booleans, objects — without manual JSON encoding on either side. Alchemy serializes at deploy and the typed runtime binding gives you the original shape back.

const Api = yield* Cloudflare.Worker("Api", {
  main: "./src/worker.ts",
  bindings: {
    MAX_ITEMS: 100,                         // number
    DEBUG: true,                            // boolean
    FEATURES: { beta: true, alpha: false }, // object
  },
});

// inside the Worker
const limit    = env.MAX_ITEMS;   // ^? number — 100
const debug    = env.DEBUG;       // ^? boolean — true
const features = env.FEATURES;    // ^? { beta: boolean; alpha: boolean }

Previously you’d round-trip through JSON.stringify at the binding site and JSON.parse on the runtime side with an as cast for typing. Now it’s just env.MAX_ITEMS.

— Thanks to Michael K (#269) for the contribution.

Neon logical replication

Opt-in flag on Neon.Project for enabling logical replication — needed for downstream CDC tools, change-data pipelines, and some replicas. Defaults to off.

const Project = yield* Neon.Project("App", {
  region: "aws-us-east-1",
  enableLogicalReplication: true,
});

Toggling the flag on an existing project applies in place — no replacement needed.

— Thanks to Baptiste Arnaud (#268) for the contribution.

CLI: version-on-npm warning

alchemy now checks the latest version on npm at startup and prints a one-line nudge when you’re behind. Skips in CI / non-TTY automatically.

$ alchemy deploy
ℹ A newer version of alchemy (2.0.0-beta.36) is available.
  Run `bun add alchemy@latest` to upgrade.

CLI: plain-text renderer in non-interactive terminals

The interactive deploy UI now downgrades to a plain-text progress renderer when run outside a TTY (CI logs, piped output). No more ANSI control sequences cluttering CI logs.

CLI: `alchemy cloudflare` / `alchemy aws` namespaces

Cloud-scoped subcommands are now grouped under provider namespaces — e.g. alchemy cloudflare state for inspecting the Cloudflare-backed state store. Top-level commands (deploy, destroy, plan, dev, test) are unchanged.

alchemy cloudflare state ls     # list state entries
alchemy aws state get <key>     # read a state entry

Fixes worth knowing about

Auth providers moved into layers. The R requirement on stack effects is never again — credential resolution is internal to the provider layer instead of leaking into the user-facing stack type.
Dev mode Node compatibility & profiles. Several Node-specific runtime issues in alchemy dev fixed, including profile handling for AWS/Cloudflare credentials.
QueueConsumer.listConsumers paginates. Previously truncated at the API’s default page size when a Worker was listed as a consumer on many queues. Now paginates and surfaces conflicting consumer registrations with a clear error.
R2 custom domain interface simplified to a plain string array (was a verbose object shape).
Drizzle schema flag tightened — only marked as updated when migrations actually drift, not on every drizzle-kit generate run.
Vite builder.sharedConfigBuild disabled. Fixes a class of build failures when bundling Workers via the Vite plugin.
Distilled Cloudflare runtime bumped to 0.3.1.

Contributors

Big thank-you to everyone who shipped code in this beta:

Alex — R2 bucket custom domains (#241)
Michael K — non-string env bindings on Workers (#269)
Baptiste Arnaud — Neon enableLogicalReplication (#268)