How to Keep Your Render Free-Tier App Alive (Without Paying for It)

If you've ever deployed a side project on Render's free tier, you've probably run into the same frustrating experience: you share the link with someone, they click it, and then... nothing happens for 30–60 seconds while the app wakes up from its sleep.

This is because Render spins down free-tier web services after 15 minutes of inactivity. The next request has to cold-boot the entire application before it can respond. For a Spring Boot app like my project ChatUp — a real-time WebSocket chat app with Groq AI — that delay kills the first impression entirely.

The good news: you can work around it for free, with very little effort.

Render's free plan is genuinely generous — you get a live HTTPS URL, automatic deploys from GitHub, and no credit card required. The trade-off is that idle services are d to conserve resources. The moment your service receives a request while sleeping, Render queues the request and starts the container. Depending on your stack, that spin-up can take anywhere from 10 seconds (lightweight Node.js) to over a minute (Spring Boot with a full application context).

The fix is simple: don't let it sleep in the first place. Send a request to your app every 10 minutes or so, and Render will never spin it down.

Here are three ways to do that.

Cron-job.org is a free service that lets you schedule HTTP requests on a cron schedule, no account upgrade required.

Steps:

https://your-app.onrender.com

)That's it. The dashboard shows you a log of every execution, so you can verify your app is returning 200 OK

. If it's returning errors, you'll know immediately.

Best for: Anyone who wants this solved in under 5 minutes with no code.

If your project is already on GitHub, you can add a workflow that pings your app on a schedule. This keeps everything in one place and doesn't require a third-party account.

Create .github/workflows/keep-alive.yml

in your repo:

name: Keep Alive

on:
  schedule:
    - cron: '*/10 * * * *'  # every 10 minutes

jobs:
  ping:
    runs-on: ubuntu-latest
    steps:
      - name: Ping Render App
        run: curl -s https://your-app.onrender.com/ > /dev/null

Push it to your default branch and GitHub Actions will handle the rest. You can verify it's running under the Actions tab in your repository.

Caveat: GitHub Actions has a known behaviour where scheduled workflows on repos with no recent activity may be d after 60 days. You can re-enable them manually, or commit something small to reset the timer.

Best for: Developers who prefer keeping infrastructure-as-code in their repo.

For Spring Boot apps specifically, you can add a scheduled task that pings the app's own health endpoint. This requires no external service.

First, enable scheduling in your main class or a config class:

@SpringBootApplication
@EnableScheduling
public class ChatUpApplication {
    public static void main(String[] args) {
        SpringApplication.run(ChatUpApplication.class, args);
    }
}

Then add a scheduler component:

@Component
public class KeepAliveScheduler {

    private static final Logger log = LoggerFactory.getLogger(KeepAliveScheduler.class);

    @Value("${app.base-url}")
    private String baseUrl;

    @Scheduled(fixedRate = 600_000) // every 10 minutes
    public void keepAlive() {
        try {
            HttpURLConnection connection = (HttpURLConnection)
                new URL(baseUrl + "/actuator/health").openConnection();
            connection.setConnectTimeout(5000);
            connection.connect();
            log.info("Keep-alive ping sent. Response: {}", connection.getResponseCode());
        } catch (Exception e) {
            log.warn("Keep-alive ping failed: {}", e.getMessage());
        }
    }
}

Set app.base-url=https://your-app.onrender.com

in your application.properties

.

Important caveat: This approach has a fundamental weakness — if the app is already asleep, it cannot ping itself to wake up. It only helps prevent the app from sleeping in the first place, which means if Render does spin it down (e.g., after a redeploy or a brief outage), the self-ping won't kick in until the app is already running again. Use this as a supplement to one of the external approaches above, not as a replacement.

Best for: Adding a redundant keep-alive layer on top of an external pinger.

For my ChatUp app, I went with cron-job.org — it took about two minutes and the app has been instantly responsive ever since. If you want belt-and-suspenders reliability, combine cron-job.org with the GitHub Actions workflow for redundancy.

If your app gets real traffic and the free tier becomes a limitation, Render's Starter plan ($7/month per service) removes the spin-down behaviour entirely. For a production app, that's worth it. But for side projects and demos, the approaches above keep things running without spending a dime.

Render's free tier is a fantastic way to get a project live quickly. The spin-down behaviour is the one real friction point, but as you've seen, it's easy to work around. Pick whichever approach matches your setup and you'll never have to apologise for a slow first load again.

If you want to check out the project this post came from, ChatUp is live at chatup-jg85.onrender.com and the source is on GitHub.