Render

Render is a cloud platform for hosting the kind of apps most small teams actually build: a web service in Node, Python, Ruby, Go, Rust or Elixir, a Postgres database behind it, a cron job or two, and a static front end. You connect a GitHub, GitLab or Bitbucket repo, pick a compute plan, and Render builds and deploys it. Push to the linked branch and it redeploys on its own.

How deployment works on Render

Auto-deploy on commit is the default for a new service. You can switch it to run only after CI checks pass, or turn it off and deploy by hand. Every service type redeploys with zero downtime unless it has a persistent disk attached. Service previews spin up a temporary copy of a web service or static site with its own onrender.com URL so you can test a pull request, and Blueprints (a render.yaml file in your repo) describe a whole stack of services and databases as code.

Free tier limits to know before you build on it

Free web services, Postgres and Key Value instances are meant for testing, and Render says not to use them for production. A free web service spins down after 15 minutes without traffic and takes about a minute to wake up. It gets 750 free instance hours a month per workspace, cannot scale past one instance, has no persistent disk or SSH access, and cannot send mail over ports 25, 465 or 587. Its filesystem is wiped on every restart or redeploy. A free Postgres database holds 1 GB, has no backups, and expires 30 days after creation, with a 14 day grace period to upgrade before it is deleted.

Scaling and production costs

Pricing has two layers. The workspace plan sets platform features: Hobby is $0 a month, Pro is $25 a month and Scale is $499 a month, with Enterprise on custom pricing. Compute is billed on top, prorated by the second. A web service on the Free plan costs nothing, a 0.5 CPU / 512 MB instance is $7 a month, 1 CPU / 2 GB is $25 a month, and 2 CPU / 4 GB is $85 a month. Managed Postgres starts at $6 a month for 256 MB, and a 1 GB instance is $19 a month. You can run multiple instances on any paid plan, up to 100 for most sizes, but horizontal autoscaling needs a Pro workspace or higher. Bandwidth is included per plan (5 GB on Hobby, 25 GB on Pro, 1 TB on Scale), and overage is billed.

The trade-offs

Render is easy to start on and the dashboard is clear about what each service costs. The price of that is that production setups add up: a paid web service, a paid database and a Pro workspace for autoscaling and unlimited seats come to well over the headline $0. Teams that want fine-grained control over networking and infrastructure will hit the limits of a managed platform sooner than they would on a hyperscaler. For indie projects, startups and small teams that want Git-to-live without managing servers, it is a sensible pick.