Rate limiting

Edifix rate-limits requests, not the number of references inside a job. A single create_job call carrying a thousand references counts as one request.

There are two independent rate limits:

Limit	Endpoints	Refill rate	Request limit
Create-job limit	POST /create_job, POST /create_job_from_file	One request every three seconds (~20/min sustained)	1200 requests
General limit	All other endpoints	One request every 0.3 seconds (~200/min sustained)	12000 requests

Capacity returns continuously at each limit's refill rate, so you don't have to wait idle after hitting a limit—you can resume immediately at the reduced rate. RateLimit-Reset reports when the full request limit is available again, about one hour from empty for either limit.

Scope

Single-session accounts are rate-limited per user.

Multi-session accounts are rate-limited per session where each session_tag has its own rate limit. See Authentication for multi-session details.

Response headers

Every response reflects the rate-limit that applied to the request (including 429), and carries the following headers:

Header	Meaning
RateLimit-Limit	The request limit for this endpoint: 1200 for create-job calls, 12000 for everything else.
RateLimit-Remaining	Requests remaining before you exceed the limit.
RateLimit-Reset	Seconds until the full request limit resets.

TIP

Instead of waiting for a 429 response, monitor RateLimit-Remaining, and reduce your request rate before reaching the limit.

Exceeding the limit

When you exceed the limit, the API returns:

• HTTP status 429 Too Many Requests
• Error code 0028 in the JSON body (see Error codes)
• The same three rate-limit headers, where RateLimit-Limit reflects the limit you exceeded (1200 for create-job calls, 12000 for all others), RateLimit-Remaining: 0, and RateLimit-Reset shows how long until the limit resets.

Example:

{
  "errors": [
    {
      "error_code": "0028",
      "error_message": "Unable to process references because your account has hit the rate limit threshold for this endpoint. See the response headers for rate limit reset details."
    }
  ]
}

Recommendations

• Polling guidance: Wait approximately two seconds per submitted reference before your first check, then check no more than once every three seconds.
• Honour RateLimit-Reset: Do not retry before RateLimit-Reset indicates capacity has returned.
• Stagger large batches: Avoid submitting thousands of jobs in quick succession — once exhausted, requests are throttled back to the refill rate for up to an hour before the full create-job limit returns.
• Adapt to limit changes: RateLimit-Limit always reports the current request limit, so your integration adapts automatically if the published limits change.