Bought a "residential" proxy or a new VPS? IP Clean Check tells you in one click whether sites will see a real home/mobile user or a datacenter. It cross-checks 2 IP databases, gives a 0–100 clean score, explains every deduction (hosting ASN, proxy/VPN/Tor flags), and runs browser leak checks (timezone, WebRTC, language). Look up any IP or call the free JSON API (batch up to 10). Built for the Cloud Run Hackathon: zero-dependency Python, deployed with one gcloud run deploy, scales to zero.

How did you use Cloud Run for this launch?

IP Clean Check is built and deployed entirely on Google Cloud Run. Cloud Run hosts the serverless service that receives a request, inspects the public IP, ASN, ISP and geolocation, and returns a quick residential-vs-datacenter verdict. Using Cloud Run lets the tool scale to zero when idle, deploy in seconds from our codebase, and stay free and accessible with no sign-up.

Hey Product Hunt 👋

I'm Cong, a technician in Shenzhen. I run a handful of small VPS and proxy servers, and I kept hitting the same question: is this IP actually "clean"? Providers sell "residential" IPs, but sites, payment providers and streaming services quietly treat datacenter IPs as high-risk. Existing checkers either give a bare score with no explanation or bury the answer under ads.

So I built IP Clean Check:

✅ Open it and you instantly see whether your IP looks Residential, Mobile or Datacenter, with a 0–100 clean score

✅ Every deduction is spelled out (e.g. "AS15169 belongs to Google, a known cloud network −35"), cross-checked against two IP databases

✅ Browser leak checks that run only on your device: timezone mismatch, WebRTC leak, languages

✅ Look up any IP, or script it with the free JSON API, including a batch endpoint for up to 10 IPs (no key, CORS on)

Try the contrast: 8.8.8.8 scores 0 (Datacenter), while a T-Mobile mobile IP and a Chunghwa Telecom home IP both score 100.

How it's built on Google Cloud Run ☁️

• One Python file, standard library only, no Dockerfile

• Shipped with a single `gcloud run deploy --source .`

• Scales to zero when idle and scales out on its own for launch-day traffic; billed per request

• New versions go out as no-traffic tagged revisions, get tested on their own URL, then traffic moves over — zero downtime

• Fun gotcha I hit: Cloud Run reserves /healthz at the edge, so the health check lives at /health

It's free, no signup, and nothing is stored. I'd love your feedback: which signal should I add next — ASN reputation history, IPv6 detail, or a checker you can run from the command line? And if you sell or buy proxies, tell me where the score feels wrong. 🙏