Probability-scored AI-infrastructure data, via REST.

Quickstart

Three steps. Should take under 60 seconds.

1. 1. Get a test key.

   Sign up for a free test key — your gc_test_… key is issued immediately and emailed to you. Test keys run against the same endpoints as live keys but cannot bill.

2. 2. Set the Authorization header.

   bash

   Copy

   export GREENCIO_API_KEY="gc_test_..."

3. 3. Make your first request.

   curlGET /v1/capacity/signals

   Copy

   curl https://www.greencio.com/api/v1/capacity/signals?limit=1 \
     -H "Authorization: Bearer $GREENCIO_API_KEY"

   You will get back an envelope like this:

   json200 OK

   Copy

   {
     "data": [
       {
         "id": "sig_01HZF8N7C3K9YR3P6Q5R2VTYBW",
         "type": "permit_filing",
         "title": "Ohio data center permit — PUCO filing detected",
         "impact": "high",
         "observed_at": "2026-05-12T14:08:00Z",
         "operator": "Meta",
         "region": "US-OH",
         "source_url": "https://puco.ohio.gov/...",
         "linked_markets": ["mkt_01HZF8K..."]
       }
     ],
     "meta": {
       "generated_at": "2026-05-16T00:00:00Z",
       "request_id": "req_01HZH5J7K4M9Q8...",
       "next_cursor": "eyJpZCI6InNpZ18wMUhaRjhONy...",
       "source_mode": "seed_reference"
     }
   }

That's it. Same shape across every endpoint — list responses always come with a next_cursor, single resources omit it. Continue with Authentication to understand key lifecycle, or jump to Reference for the endpoint list.

Authentication

Every request must include an Authorization header with a Bearer token. Keys come in two flavors:

curl https://www.greencio.com/api/v1/index/compute \
  -H "Authorization: Bearer $GREENCIO_API_KEY"

Key lifecycle

• Issued: shown once at creation time. Store it immediately.
• Active: validates on every request. last_used_at updates each call.
• Revoked: returns 401 unauthenticated. Rotation is hand-driven today; a self-serve rotation UI lands in v1.1.

Conventions

Every endpoint in the API follows the same conventions. Once you know them you know all of v1.

Response envelope

Success responses always have data and meta. Error responses always have error. Errors never appear inside data.

{
  "data": [ /* items */ ],
  "meta": {
    "generated_at": "2026-05-16T00:00:00Z",
    "request_id": "req_01HZH...",
    "next_cursor": "eyJpZCI6...",
    "source_mode": "seed_reference"
  }
}

{
  "error": {
    "code": "invalid_filter",
    "message": "since must be ISO-8601 UTC",
    "field": "since",
    "request_id": "req_01HZH..."
  }
}

The request_id appears in both the body and the X-Request-Id response header. Include it in any support email and we can trace the exact call.

Pagination

Collections are cursor-paginated. There is no offset pagination and there will never be — it does not survive data drift.

# First page
curl "https://www.greencio.com/api/v1/capacity/signals?limit=50" \
  -H "Authorization: Bearer $GREENCIO_API_KEY"

# Next page — pass meta.next_cursor from the previous response
curl "https://www.greencio.com/api/v1/capacity/signals?limit=50&cursor=eyJpZCI6..." \
  -H "Authorization: Bearer $GREENCIO_API_KEY"

• limit defaults to 50, max 500.
• cursor is opaque base64. Don't parse it; round-trip what we sent.
• meta.next_cursor is null when the result set is exhausted.

Time format

Every timestamp is an ISO-8601 string in UTC: 2026-05-16T00:00:00Z. No Unix epoch, no locale variation. Filters that accept a time accept the same format.

Resource IDs

IDs are opaque, stable, and prefixed by resource type. Customers store them; the prefix prevents the “is this a signal or a facility ID?” support load.

Errors

Every error response uses the same shape and one of a fixed taxonomy of codes. Match on code, not on message — messages may change for clarity, codes will not.

Rate limits

Two independent limits: a per-second burst and a monthly quota. Both are surfaced on every response, including success responses, so you don't need to hit 429 to learn what's left.

X-RateLimit-Limit: 30
X-RateLimit-Remaining: 28
X-RateLimit-Reset: 2026-05-16T00:00:01Z
X-RateLimit-Quota-Limit: 50000
X-RateLimit-Quota-Remaining: 47233
X-RateLimit-Quota-Reset: 2026-06-01T00:00:00Z

When you exceed either limit you get a 429 with a Retry-After header in seconds.

Test keys run at Pro-level limits so you can iterate without throttling, but their monthly counter is separate from any live key on the same account.

Idempotency

Every POST endpoint accepts an Idempotency-Key header. It is required for endpoints that compute billing-relevant results — currently /lcoe and /inference/unit-economics. Optional elsewhere. Without it, a network retry would create a duplicate.

curl -X POST https://www.greencio.com/api/v1/lcoe \
  -H "Authorization: Bearer $GREENCIO_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: $(uuidgen)" \
  -d '{
    "capex_usd": 1200000000,
    "annual_generation_mwh": 2400000,
    "project_lifetime_years": 20,
    "discount_rate": 0.08,
    "annual_opex_usd": 40000000
  }'

• Send the same key with the same body → original response is replayed. Safe to retry.
• Send the same key with a different body → 409 idempotency_conflict. Pick a new key.
• Keys are valid for 24 hours. After that, the slot is reusable.
• Use UUID v4, a request hash, or any 8–200 char string. We don't parse it; we match it.

Reference

Five products. Endpoint tier indicates which subscription plan grants access — see Pricing.

curl "https://www.greencio.com/api/v1/capacity/signals?impact=high&operator=meta&limit=2" \
  -H "Authorization: Bearer $GREENCIO_API_KEY"

curl https://www.greencio.com/api/v1/index/compute \
  -H "Authorization: Bearer $GREENCIO_API_KEY"

curl -X POST https://www.greencio.com/api/v1/inference/unit-economics \
  -H "Authorization: Bearer $GREENCIO_API_KEY" \
  -H "Content-Type: application/json" \
  -H "Idempotency-Key: ue-2026-05-16-001" \
  -d '{
    "facility_capacity_mw": 120,
    "gpu_count": 8192,
    "gpu_type": "H100",
    "capex_usd": 2500000000,
    "power_cost_usd_per_mwh": 68,
    "utilization_rate": 0.72
  }'

Pricing

Pricing below is a hypothesis we are validating with the first design partners. Don't expect it to be on a buy-now button quite yet — talk to us if you want to lock in.

Enterprise plans (custom SLAs, dedicated support, EU data residency on roadmap) are negotiated. Email hello@greencio.com.

Status & changelog

• OpenAPI spec: docs/api/openapi.yaml in the repo. The contract.
• Have a regression or a question? Email hello@greencio.com with the failing request's request_id.

Changelog

• v1.0.0 — initial release. Capacity Signals, Compute Index, Inference Economics, Fabs, LCOE. Power Forecast and Carbon Trail namespaces reserved for v1.1 / v1.2.