nnesly

Your code is live minutes after you push it — and one command undoes a bad release.

Connect a GitHub repository or run one command from your project folder. nesly detects the stack (Node.js, Python, Go, .NET, Ruby, PHP, Java, Rust, a static site, or your own Dockerfile), builds it, checks that it answers, and prints the URL. There are no servers to set up, patch or watch.

terminal
$ nesly deploy --yes
[queued] deployment accepted
[building] building .NET with the official SDK image for the framework this project targets
[deploying] health check: app answered (HTTP 200)
[ready] serving at https://my-app-x7k2.nesly.net
$ nesly rollback
Rolled back to dep_3k2m
URL: https://my-app-x7k2.nesly.net

How it works

  1. 1

    Connect or run

    Connect the GitHub repository so every push deploys, or run nesly deploy from the folder. The first run creates the project and links the folder to it.

  2. 2

    We build and check

    The stack is detected and built with its official image; a Dockerfile in the root wins. The release goes live only after your app answers a health check.

  3. 3

    Share the URL

    Production serves at your project's address and your own domains. A preview deploy gets an address of its own and leaves the live site untouched.

What you get

A preview for every change

nesly deploy --preview publishes to its own address; --alias staging keeps that address stable. Pull requests on a connected repository get a preview each, with an empty database of their own.

Rollback is one command

nesly rollback moves production back to the previous release without building again; nesly promote <id> makes any earlier deployment the live one.

Logs that follow the request

nesly logs --follow streams build and runtime lines; every response carries a request id you can search for. nesly diagnose names the failure and its fix in a sentence.

Nothing to run idle

On Free an app that gets no traffic pauses, and the next request wakes it; you pay nothing for the quiet hours. Pro keeps it warm so no visitor waits.

Isolated by default

Each application runs in its own sandbox with a network policy that lets out only what you allow; secrets are stored encrypted and arrive as environment variables.

Your own domain in a minute

Add a domain, point it at us, and HTTPS is issued and renewed for you. Pin one address to one release when a client must see exactly that version.

Pricing

Free is for trying and small projects: 3 projects, 100 build minutes and 50 GB of traffic a month. Pro is TRY 869 a month with VAT in Turkey, or $18 without VAT elsewhere.

  • Pro: 20 projects, 1,000 build minutes, 500 GB of traffic, 2,048 MB of memory per process, always-on applications and custom domains.
  • Past the included amounts Pro is metered at the published unit prices, capped by a spend limit you set; your sites keep serving at the cap.
See plans and allowances

Use cases

A product built with an AI assistant

Your assistant writes the code and runs nesly deploy; you review the preview URL and promote it. No infrastructure vocabulary needed on either side.

Client sites for an agency

One project per client, a preview address for sign-off, their own domain at launch, and a rollback when a change lands badly.

An API with a database

Node, Python, Go or .NET behind HTTPS, a managed database attached with one command, backups kept nightly.

Questions, answered

Which languages and frameworks run here?
Node.js, Python, Go, .NET, Ruby, PHP, Java and Rust are detected and built with their official images. Frontends (Vite, React, Astro, SvelteKit, Angular, Next.js) and static site generators (Hugo, Jekyll, MkDocs) are built and served as files. A Dockerfile in the root is built exactly as written.
Why is the first request slow after my app has been idle?
On Free, an application that receives no traffic pauses so you pay nothing for it. The first request wakes it and takes a few seconds. Pro keeps your application always on.
Can I keep a file on disk?
The container's disk is temporary and is gone on the next deploy. Store files in an object-storage bucket and data in a managed database; both are one command away and arrive as environment variables.
What happens when a deploy fails?
The live site is untouched; production only changes once the new release answers its health check. nesly diagnose reads the failed deployment and says what to change.

Related products

Deploy something in the next five minutes

Create an account, run one command in your project folder, share the URL.