Cloudflare Buys Deno and Will Wind Down the Runtime in a Year

Cloudflare is buying all of Deno. The company will support the Deno runtime for one more year with monthly releases that carry bug fixes and security updates. After that, Cloudflare will stop developing it. The code will stay open source, and the company openly invites anyone who wants to carry the project forward. The deal was announced on 9 October 2026. By the parties' own statements, the main asset in it is not the runtime itself but celld, which the Deno team released in August.

What Cloudflare is actually getting

celld is an open-source implementation of the Durable Objects pattern from Cloudflare Workers. According to Ryan Dahl, the creator of Deno and Node.js, the system works "surprisingly well" and relies only on object storage for coordination and persistence. He describes it as a new model of server-side development, not another API for the file system or the network.

Cloudflare put its goal this way: make self-hosting workerd a fully supported way to build and run applications on the Workers programming model. For a platform whose model has so far lived mainly inside its own infrastructure, this is a step toward portability. The Deno team has already built a working open-source equivalent, which settles the question faster than building it internally from scratch.

Why the runtime will not continue

On Hacker News, Dahl called the decision a joint one and said he agrees with it. In his assessment, Deno is well designed and contains good ideas, but it does not solve truly big problems. The project got pulled into the "gravity well" of Node compatibility, which forces it to behave exactly like Node. His question is blunt: why rewrite Node if Node works? Small gains in performance, usability or security are not enough to justify that.

This is rare candor from a project's author: he himself admits that Deno's main value is moving out of the runtime and into a different architectural idea.

Key facts of the deal:

• Format: Cloudflare is acquiring Deno in full.
• Runtime support: 12 months with monthly releases.
• Release contents: bug fixes and security updates; no new features are promised.
• License: Deno remains an open-source project, and forks and community-led continuation are welcome.
• Main asset: celld, whose first version came out in August 2026.
• For comparison: permissions appeared in Node.js in v20.0.0 in April 2023 and became stable in v22.13.0 in January 2025.

Permissions: what gets lost

Simon Willison, the author of the post, names Deno's permission system as his favorite feature. A script runs with an exact list of files and folders it can read and write, and of network hosts it is allowed to reach. Node.js has a similar model, but with an important limitation: it does not support allowlisting specific hosts. Network access is either on or off.

For anyone running untrusted code, including model-generated code, this difference matters. Granular control over network access remains an advantage of Deno, and once the runtime winds down, you will have to either find it in forks or cover it with other isolation measures.

What teams on Deno should do

A year of support gives you a reasonable window for planning, but it is no reason to put off the decision. Check how much your code depends on Deno-specific APIs, as opposed to common web standards and Node compatibility. The higher the compatibility, the cheaper the migration will be. If host-level permissions are critical for you, evaluate alternatives in advance: container-level or process-level sandboxing, or a future fork.

It is also worth watching celld and how Cloudflare formalizes support for self-hosting workerd. If the promise comes true, the Workers model could run outside Cloudflare's cloud. For teams worried about vendor lock-in, that will be a weighty argument.