Skip to main content
Limits apply per organization. Every API key and OAuth token in an organization, including keys used through the MCP server, draws on the same budgets. A request over a rate limit returns 429 with a Retry-After header, and nothing runs.

How limits are counted

A request counts once Sixtyfour accepts it, even if it fails afterwards. Each endpoint has its own budget unless a table below says it shares one. A /v1 route has the same limit as the original route it replaces. The tables list both, with the /v1 route first.

Status polling

At one poll every 10 seconds, the budget covers 50 Jobs in flight, with the polls spread out rather than sent together. For more work than that, receive results by webhook and poll only as a fallback: pass webhook_url when you start a Job, and add an outgoing_webhook block to Workflows. See Outgoing Webhooks. A 429 on a status read does not affect the Job; read again after Retry-After. Async Jobs has a polling helper.

Limits by endpoint

Files

File download URLs use the existing file-download budget. Two uploads can run concurrently per organization. Multipart uploads accept up to 45 MiB; direct upload tickets accept up to 100,000,000 bytes and 100,000 data rows.

Jobs and exports

POST /v1/exports allows 20 requests per minute per organization, including replays, and two simultaneous submissions. Export rendering continues asynchronously after submission.

Intelligence and contact

Workflows

Monitors

Atlas

Webhooks

Feedback

Team skills

Connections

Media

OAuth tokens

POST /oauth/token issues up to 60 tokens per minute per client. Requests that fail client authentication are also limited to 25 per minute per client and to 120 per minute shared with other callers. Reuse each token until shortly before its expires_in. See Using OAuth Clients.

Concurrency limits

Each organization runs up to 5 deep searches at a time, people and company combined. A start above that is never refused with 429. It returns 200 with status: "running", and the search then ends failed with this error:
Start a new deep search only after one of yours reaches a final status.

History exports

Each organization runs up to 5 history-export builds at a time. A sixth returns 429 with Retry-After: 5 and the detail Five history exports are already building for this organization. Retry later.

Response headers

Successful responses carry no X-RateLimit-* headers. A 429 from the status polling budget:
A 429 from a per-minute limit:
The original routes return the same detail with code and correlation_id in place of error_code and request_id.

Retry after a 429

  • Wait at least Retry-After seconds, then send the same request again.
  • Workers that share an organization share its budgets. Coordinate them, and add a small random delay after the server’s minimum.
  • Stop after a bounded number of attempts.
  • For many records, upload a file and start one Job, or use Workflows, instead of sending one request per record.
This helper retries 429 and 503 responses after Retry-After:
Contact support@sixtyfour.ai if a limit blocks your integration. For other status codes, see Handling Errors.