Rate limits
Three policies, and the headers that show your remaining capacity.
Limits apply to your organization, for all of its keys together. More keys do not give more capacity.
Test mode has the same limits, with separate counters.
The policies
| Policy | Limit | Endpoints |
|---|---|---|
sustained | 20 req/s | All |
geocoding | 10 req/s | POST /v1/orders, POST /v1/orders/quote, POST /v1/addresses/validate |
order-list | 2 req/s | GET /v1/orders |
Univooz measures over a 10-second period. So sustained permits 200 requests in 10
seconds. Short groups of requests are acceptable. A continuous loop is not.
The narrow policies are additional to sustained. A GET /v1/orders uses both.
Read the headers
Every response gives the policies that apply and your remaining capacity:
RateLimit-Policy: "sustained";q=200;w=10, "order-list";q=20;w=10
RateLimit: "sustained";r=196;t=7, "order-list";r=18;t=7| Key | Meaning |
|---|---|
q | The total permitted |
w | The period, in seconds |
r | Your remaining requests |
t | Seconds until the reset |
Watch r and slow down before Univooz rejects you.
If you go over
429 rate_limit_exceeded, with Retry-After in seconds.
The message gives the policy.
Obey Retry-After. An earlier attempt uses capacity you do not have.
How often to read the order list
One time every 30 seconds.
A plan of 5,000 orders each month is approximately seven orders each hour. One request each second gives approximately 500 requests for each order.
GET /v1/orders has the lowest limit on the API, because it is the status channel.
For faster information, use webhooks.