Hi Product Hunt 👋 I kept watching people buy rural land the same way: see it once, on a sunny day, with the seller standing next to them. Then find out about the wetland, or the shrink-swell soil, or the fact that the parcel sits inside a National Monument. None of that information is secret. It's public, free, and scattered across a dozen federal agencies that don't talk to each other. Checking all of them properly takes about a week, so almost nobody does. LandBenchmark does it in about a minute. Draw a parcel anywhere on Earth and it measures 55 signals — flood position, wetlands, soil productivity, depth to bedrock, slope, protected areas, contamination records, access. 34 work anywhere in the world; 21 are US-only, where the public data runs deepest. The part I actually want feedback on: Every value names the agency it came from, the date it was observed, and links to the source. And we measured our own error rate and published it — including the parts we've never checked, which say UNMEASURED with the reason. Our elevation is bounded against 22 NGS survey marks (mean absolute error 0.904 m); climate sits at 0.674 °C against NOAA station normals. It's public, no signup: landbenchmark.com/accuracy Doing that measurement found a real bug in our own product — a corrupt tile in a federal dataset was silently downgrading US parcels to a coarser global model, and no test caught it because the number still looked plausible. For the builders here: there's a REST API and an MCP server, so you can run this from Claude Desktop, Claude Code, or your own agent. → /developers Free: one full report per account, no card. Pro is $19/mo with a 3-day trial. Here's a complete report on a real parcel, nothing gated: landbenchmark.com/sample-report What I'd genuinely like to know: if you've bought land, what did you wish you'd known before you signed? I'm here all day.

LandBenchmark

Check any parcel before you buy — with sources

Check any parcel before you buy — with sources