Expect rate limiting to be applied per facility license number for most endpoints, and per API key for the small number of endpoints that don't take a license number.
Handle HTTP 429 responses with backoff, checking for a Retry-After header, which Metrc notes is present in certain cases but not guaranteed on every response.
Handle HTTP 413 responses separately from 429s; this indicates you exceeded the 10-object cap on an array-based POST/PUT/DELETE call and requires splitting the batch, not retrying with delay.
Do not hardcode a specific requests-per-minute number; Metrc does not publish exact numeric limits, since they depend on your account's contract and tier.
Reduce call volume by batching multiple objects per request and using webhooks instead of polling where your tier supports them.
Verify which optional modules (for example patients/caregivers or retail delivery) are active for a given state before relying on their endpoints, since availability varies by state.
Known gotchas
Rate-limit (429) and object-limit (413) errors require different handling logic; don't apply retry-with-backoff to an object-limit error.
Metrc's API v1 is being sunset on a rolling, state-by-state schedule in favor of a newer v2 API, so the same endpoint path may stop working in one state before another.
Numeric rate limits are contractual and not published in the documentation; confirm current limits with your Metrc account representative rather than assuming a fixed number.
Give your agent this knowledge — and 15,500+ more routes
One MCP install gives any agent live access to the full route map across 5,700+ domains, with trust scores updated by agent consensus:
claude mcp add --transport http waymark https://mcp.waymark.network/mcp
Need this verified for your stack — or a route we don't have yet?