API and MCP server

Plug GridCensus into your own models and AI tools. The MCP server answers anyone in public mode; full records, the REST API, the change feed, webhooks and weekly data files come with Enterprise.

MCP server (AI tools)

Connect GridCensus to Claude, ChatGPT or your own agent. The server is https://gridcensus.com/api/mcp (Streamable HTTP). Without a key it answers in public mode: summaries with a link to each record, 5 results per call. With an Enterprise key it returns full records, change history and the change feed.

{
  "mcpServers": {
    "gridcensus": {
      "type": "http",
      "url": "https://gridcensus.com/api/mcp",
      "headers": { "Authorization": "Bearer gck_live_…" }
    }
  }
}

Tools: search_sites, get_sites, search_datacenters, get_datacenters, search_queue-projects, get_queue-projects, search_brownfields, get_brownfields, recent_changes (Enterprise), about.

REST API (Enterprise)

Create a key under Account → API keys, then send it as Authorization: Bearer gck_live_…. JSON responses; keyset paging with cursor; up to 100 records per page and 120 requests a minute per key.

curl -H "Authorization: Bearer $KEY" \
  "https://gridcensus.com/api/v1/sites?state=TX&limit=100"

Endpoint	What it returns
`GET /api/v1`	datasets, filters, your limits
`GET /api/v1/sites`	Scored datacenter candidate sites: DC Readiness score, nearest substation and capacity, fiber, water, hazard, utility, queue activity and bottlenecks nearby. Filters: `statesite_typeiso_regionfips_code` · `q` = name prefix
`GET /api/v1/datacenters`	Existing and announced US datacenters: operator, location, capacity, square footage, type and status. Filters: `stateoperator` · `q` = name prefix
`GET /api/v1/queue-projects`	Active ISO interconnection-queue requests (generation and storage): ISO, queue ID, MW, fuel, status, queue date, county. Filters: `statefipsiso`
`GET /api/v1/brownfields`	Retired power plants and EPA brownfields with existing grid connections: capacity, voltage, retirement date, cleanup status. Filters: `statesite_type` · `q` = name prefix
`GET /api/v1/sites/{id}`	one candidate site + its last 20 changes
`GET /api/v1/datacenters/{id}`	one datacenter + its last 20 changes
`GET /api/v1/queue-projects/{id}`	one interconnection-queue project + its last 20 changes
`GET /api/v1/brownfields/{id}`	one brownfield site + its last 20 changes
`GET /api/v1/changes`	change feed: new, updated, removed records (`dataset`, `after_id`, `since`)

Bulk files and webhooks

Enterprise includes every national and state file (gzipped CSV, refreshed weekly), and any alert can post to your webhook, signed with X-Census-Signature: t=…,v1=HMAC-SHA256(secret, t.body).