API reference

A compact map of the surface. All responses are JSON; all authenticated endpoints take a Bearer key.

Comps

GET /v1/scrape
Sold (or active) listings for a keyword, cleaned and summarized.
GET /v1/sold
Alias for a sold-only scrape.
POST /v1/scrape/bulk
Up to 25 queries in one call, each cache-first and metered once.
POST /v1/scrape/jobs
Enqueue bulk/bursty work; returns 202 + job_id.
GET /v1/scrape/jobs/{job_id}
Poll an async job.

Intelligence

GET /api/research
Grade-by-grade single-card research (raw/PSA 9/PSA 10), honest-by-construction.
POST /v1/shipping/rates
Observed shipping for a keyword + a labelled landed-cost estimate.

Account

GET /v1/status
Service status, limits and capabilities.
GET /v1/usage
Current key usage, plan, quotas and reset window.
GET /v1/ledger/stats
Corpus size and cache stats.
GET /api/account/keys
List your API keys (dashboard; session-authenticated).
POST /api/account/keys
Create an API key.
POST /api/account/keys/{id}/rotate
Rotate a key.
POST /api/account/keys/{id}/revoke
Revoke a key immediately.

Machine-readable schema

The full OpenAPI 3.1 document is served at /openapi.json — use it to generate typed clients (openapi-generator, orval, etc.). A live reference UI is not published yet; the schema is the source of truth.

Response headers

Every metered /v1 response tells you exactly where you stand — read these instead of guessing:

Header	Meaning
`X-Usage-Limit`	Monthly request allowance on the current plan.
`X-Usage-Remaining`	Requests left this calendar month.
`X-RateLimit-Limit`	Allowed requests per minute (or “unlimited”).
`X-RateLimit-Remaining`	Requests left in the current minute window.
`X-RateLimit-Reset`	Seconds until the per-minute window resets.
`X-Cache`	Whether this call was served from corpus or went live.

Pagination

A standard call returns up to 240 cleaned rows in one response — enough for most comps. When you need the deep tail, use Max Mode (async paging) instead of looping page after page.

Errors & retries

Stable error codes and the retry policy live on the errors page. The rule: retry 5xx with backoff + jitter; never retry 400 or 401.

Full parameter list for /v1/scrape is on the overview page.