Rate Limits
The Revi API applies request protection to keep integrations stable.
Current Limits
| Context | Limit |
|---|---|
| Requests without an API key | 60 requests per minute per IP |
| Requests with an API key | 200 requests per minute per API key |
These values reflect the current public API limits and may change by environment.
For /api/2026-10, limits are 200 requests per minute per credential and 1,000 requests per minute per business. Both limits apply equally to every subscription plan.
These versions also protect the operations that consume the most database work:
| Workload | Per credential | Per business |
|---|---|---|
| Order/product/suppression write items | 500 items/minute | 1,000 items/minute |
| Statistics and product searches | 60 requests/minute | 240 requests/minute |
A request can contain at most 100 orders or products, its body is limited to 2 MiB, and numbered pagination accepts pages from 1 to 1,000. These workload controls are independent of the subscription plan and do not replace the general request limits. For 2026-10 product writes, the product itself, each locale, each combination and each entry in catalog_variants.variants count as one write item each. Each product can have up to 100 entries in each of those lists.
Response Headers
Successful and failed requests can include:
X-RateLimit-Limit: 200
X-RateLimit-Remaining: 193
X-RateLimit-Business-Limit: 1000
X-RateLimit-Business-Remaining: 947When the limit is exceeded:
HTTP/1.1 429 Too Many Requests
Retry-After: 42
X-RateLimit-Limit: 200
X-RateLimit-Remaining: 0
X-RateLimit-Workload: expensive_reads
X-RateLimit-Workload-Limit: 60
X-RateLimit-Workload-Remaining: 0{
"success": false,
"error": {
"code": "rate_limited",
"message": "Rate limit exceeded."
},
"meta": {
"api_version": "2026-10",
"retry_after": 42,
"limited_bucket": "expensive_reads"
}
}Client Guidance
- Use exponential backoff for transient failures.
- Stop sending requests until
Retry-Afterhas passed after a429. - Batch orders and products using the documented bulk endpoints.
- Keep each batch at
100items or fewer and split larger initial imports into several requests. - Avoid syncing all historical data on every cron run.