FraudGuard enforces tier-based limits to keep API performance predictable for every customer. Integrations should respect both the short-term request rate and the monthly request allocation attached to the account.

Limits by plan

Customer plan	Monthly API requests	Maximum request rate
Starter	1,000,000	2 requests per second
Professional	5,000,000	5 requests per second
Business	25,000,000	10 requests per second
Enterprise	100,000,000	40 requests per second

The per-second value is a throughput ceiling, not a target. Sustaining the maximum rate continuously can exhaust the monthly allocation well before the billing period ends.

Specialized products or negotiated Enterprise agreements may have additional endpoint-specific limits. The authenticated account, applicable endpoint documentation, and customer agreement are authoritative.

Build a rate-limit-safe client

Pace requests

Use a token bucket or another client-side limiter to keep bursts within the plan’s request-per-second ceiling. If multiple workers share one FraudGuard account, coordinate their combined traffic rather than limiting each worker independently.

Handle 429 Too Many Requests

Treat HTTP 429 as a temporary capacity signal:

1. stop sending immediate retries;
2. honor Retry-After when the response provides it;
3. otherwise use exponential backoff with jitter; and
4. cap retry attempts so downstream latency remains predictable.

Cache according to the response

Avoid repeating an identical lookup inside the recommended cache window. ACE v2 provides cache guidance based on the freshness and volatility of the underlying evidence. Higher-risk or rapidly changing signals may require a shorter cache lifetime.

Use bulk endpoints when appropriate

If your plan includes bulk intelligence endpoints, batch work instead of issuing a large number of individual lookups. Bulk requests reduce connection overhead, but their endpoint-specific limits and payload constraints still apply.

Monitor usage before it becomes an outage

Track request volume, 429 responses, retry counts, and remaining monthly capacity. Alert before the allocation is exhausted, especially when traffic is seasonal or driven by abuse spikes.

Review the FraudGuard API documentation for endpoint behavior and the current plan comparison for commercial details. For account-specific capacity questions, contact hello@fraudguard.io.