On October 9, 2026, Ryan Dahl, who created both Node.js and Deno, announced that the entire Deno team is joining Cloudflare. Cloudflare published a companion post by Dahl and Cloudflare's Kenton Varda under its 'Acquisitions' tag. Neither company disclosed a price or deal structure. This is more than a talent hire, because both posts set end dates for Deno's own products. The Deno runtime gets one more year of monthly bug-fix and security releases, and then the Deno team stops developing it. Deno Deploy, the company's hosting service, runs for six more months and then shuts down.
Builders are affected in two ways. Anyone running on Deno Deploy, or on a platform built on the Deno runtime, now has a migration deadline. And the deal shows where Cloudflare is heading: it wants the Workers programming model, including Durable Objects, to run on infrastructure customers own, using Deno's open-source celld daemon as the starting point.
What we checked
- What concrete evidence would resolve the remaining editorial uncertainty for this item? confirmed. Primary sources settle it. Deno's blog post by Ryan Dahl and a Cloudflare blog post by Dahl and Kenton Varda, tagged 'Acquisitions', both dated 2026-10-09, confirm the team is joining Cloudflare and lay out timelines for the runtime, Deploy, JSR and rusty_v8. Neither post discloses deal terms. [1][2]
What we found
- Deno's announcement sets the timelines. The runtime gets monthly bug-fix and security releases for one year. After that, the post says, 'we will end our development of the Deno runtime', and others are invited to continue it as open source. Deno Deploy keeps running for six months before shutting down, with migration support to Cloudflare Workers for paying customers. JSR keeps operating, with its infrastructure moving to Cloudflare. rusty_v8 support continues, with work toward integrating it into workerd. The post lists Fresh among Deno's projects but says nothing about its future. [1]
- The Cloudflare post, co-written by Dahl and Varda, is about self-hosting, not about the Deno runtime: 'we're merging workerd and celld!' Dahl and Bert Belder will lead an effort to make self-hosting workerd a first-class option. Varda acknowledges that workerd today supports Durable Objects only 'in a single-instance way'. The post gives no price or deal structure. [2]
- celld is an Apache-2.0 daemon that runs a Workers app from an existing wrangler.json. It covers Durable Objects, KV, Queues, D1, R2, Workflows, Cron Triggers and static assets. Each Durable Object is a 'cell' with its own SQLite database, and long-term state goes to S3-compatible, GCS or Azure Blob storage. Ownership of each cell is coordinated through conditional writes to the storage bucket. The repo marks containers as experimental, supports only a subset of Wrangler config, and declares products that depend on the Cloudflare network, GPUs or browser farms out of scope. [3]
What it means
If you are on Deno Deploy, start migrating now. The service has about six months from October 9, which lands around early April 2026 by the announced window, and Cloudflare Workers is the only migration path offered. If you run the Deno runtime yourself or through Supabase or Netlify, nothing breaks immediately. But upstream development ends after roughly a year of maintenance releases, and security patches after that depend on whoever forks the project, if anyone does. New projects should target Node-compatible or Workers-compatible APIs and avoid Deno-only features such as host-level network permissions. Node now has a stable permission model, but it cannot allow-list specific network hosts. Keep packages on JSR only if you're comfortable with Cloudflare running it.
The bigger signal is that Cloudflare is turning the Workers model, and Durable Objects in particular, into something you can self-host. celld already runs wrangler.json apps against plain object storage. If you've avoided Durable Objects because of lock-in, the workerd/celld merge is worth tracking. Treat celld as pre-production for now: containers are experimental, Wrangler support is partial, and the repo takes contributions only by emailed patch. Cloudflare has said only that more announcements will come 'in the coming months'.
Still open
- What are the deal's price, structure and closing status? Neither company has disclosed them.
- Will anyone credibly fork and maintain the Deno runtime once Cloudflare-supported releases end?
- When will the merged workerd/celld self-hosting release ship, and with which Durable Objects guarantees?
- What happens to Fresh, which the announcement doesn't address?