The LandBenchmark API

Authentication

Create a key on your account page. Pass it as Authorization: Bearer <key> (or x-api-key). Keys are shown once — store them securely.

Analyze a parcel

POST /api/v1/analyze — body takes a GeoJSON geometry (Polygon or Point) or a lat+lon, plus optional label and mode (lite, the default fast scan, or full for the deep satellite report).

curl -X POST https://www.landbenchmark.com/api/v1/analyze \
  -H "Authorization: Bearer tk_live_..." \
  -H "content-type: application/json" \
  -d '{ "lat": 36.02, "lon": -119.65, "mode": "lite" }'

Response

{
  "id": "…",
  "verdict": "CAUTION",
  "mode": "lite",
  "developed": false,
  "signals": [
    {
      "key": "flood-zone",
      "label": "FEMA flood zone",
      "value": "AE",
      "flag": true,
      "flag_reason": "In a FEMA Special Flood Hazard Area",
      "method": "FEMA NFHL …",
      "citation": "FEMA National Flood Hazard Layer",
      "source_url": "https://…",
      "observed_on": "…"
    }
  ],
  "unavailable": [],
  "methodology_url": "/methodology",
  "disclaimer": "Informational only …"
}

Every signal carries its method, citation, source URL, and observation date — the same evidence chain shown in the app.

mode in the response is always the mode we actually ran. If you request full and your account is not entitled to it, you receive the lite scan along with mode_requested and mode_downgraded_reason — we will not quietly hand you something other than what you asked for.

Use it from an AI agent (MCP)

LandBenchmark ships a Model Context Protocol server that exposes an analyze_parcel tool. Point any MCP-capable agent (Claude Desktop, Claude Code, custom GPTs, agent frameworks) at it with your API key, and the agent can pull a land verdict mid-conversation. See terrain-mcp/ in the repo.

{
  "mcpServers": {
    "terrain": {
      "command": "npx",
      "args": ["-y", "terrain-mcp"],
      "env": {
        "TERRAIN_API_KEY": "tk_live_...",
        "TERRAIN_BASE_URL": "https://www.landbenchmark.com"
      }
    }
  }
}