Serverless platforms are excellent for websites, but some workloads need a server that is always running: a REST API with persistent connections, a queue worker processing jobs, or a WebSocket server for real-time features. Render makes these simple to deploy, with Git-based deployments, managed Postgres and zero server maintenance.
1. When we choose Render
- Backend APIs built with Node.js (Express, NestJS, Fastify), Python (FastAPI, Django), Go or Docker.
- Background workers for emails, AI processing, imports and other queued jobs.
- Cron jobs and scheduled tasks that run longer than a serverless function allows.
- WebSocket and real-time servers.
- Managed Postgres and Key Value (Redis-compatible) stores alongside your services.
“Vercel for the frontend, Render for the long-running backend. That split covers most of the products we build.”— Strategic Takeaway
2. Prepare your app
Render tells your service which port to use through the PORT environment variable. Add a health check route so Render only sends traffic to healthy instances and can restart failing ones.
import express from "express";
const app = express();
app.use(express.json());
app.get("/health", (_req, res) => res.json({ status: "ok" }));
app.get("/api/projects", async (_req, res) => {
// ...read from the database
res.json([]);
});
const port = Number(process.env.PORT) || 3000;
app.listen(port, () => console.log(`API running on port ${port}`));3. Describe your stack with a Blueprint (render.yaml)
Instead of clicking through the dashboard for each service, we commit a render.yaml file to the repository. It creates the API, the worker and the database, and connects them automatically.
services:
- type: web
name: client-api
runtime: node
plan: starter
buildCommand: npm ci && npm run build
startCommand: npm start
healthCheckPath: /health
envVars:
- key: NODE_ENV
value: production
- key: DATABASE_URL
fromDatabase:
name: client-db
property: connectionString
- key: OPENAI_API_KEY
sync: false # you'll be asked to enter the secret in the dashboard
- type: worker
name: client-jobs
runtime: node
plan: starter
buildCommand: npm ci && npm run build
startCommand: node dist/worker.js
envVars:
- key: DATABASE_URL
fromDatabase:
name: client-db
property: connectionString
databases:
- name: client-db
plan: basic-256mbgit add render.yaml
git commit -m "Add Render blueprint"
git push
# Then in Render: New → Blueprint → select the repository4. Pricing
Render charges a workspace plan plus a monthly price for each service you run. Pricing (checked November 2026): the Hobby workspace is free, Pro is $25/month flat and Scale is $499/month flat, with Enterprise custom. Web services and workers start at $7/month for a Starter instance (512 MB RAM, 0.5 CPU), Standard is $25/month (2 GB, 1 CPU) and Pro is $85/month (4 GB, 2 CPU). Postgres starts at around $6/month.
Free web services spin down after about 15 minutes of inactivity and take a while to wake up, so they suit demos but not production. A small production backend (one Starter API, one Starter worker and a basic database) usually starts at roughly $20/month before bandwidth.
5. Production tips
- Keep the API, worker and database in the same region for low latency and free private networking.
- Use the internal service URL for service-to-service calls instead of the public internet.
- Run database migrations in a pre-deploy command so the app never starts with an old schema.
- Enable daily backups on paid Postgres plans and test a restore.
- Add autoscaling (on supported plans) for APIs with variable traffic.
- Send logs to your monitoring tool and set alerts for failed deploys and health checks.
Official resources
Principles for deliberate execution
- 1Render runs always-on services: APIs, WebSocket servers, background workers and cron jobs.
- 2A render.yaml Blueprint describes your whole stack as code, so it can be recreated in one click.
- 3Free web services sleep when idle, so use at least the Starter instance in production.
- 4You pay a workspace plan plus a price per service, so add up every service when budgeting.