Rate limits and pagination

The standard limit is 1,000 requests per minute, per user. Page through lists with limit and offset, or with cursors on the accounts list.

StartGET /accounts/?limit=50&sort_by=created&sort_order=desc
ReturnsPage of accounts
CheckMore?

YesGET /accounts/?limit=50&sort_by=created&sort_order=desc&after_id=<last id>
NoDone

Over limit: 429 + Retry-After.

Page through a list

Set your base URL and API key.

export CORIS_BASE_URL="https://sandbox-api.coris.ai/api/v0_2"
export CORIS_API_KEY="YOUR_API_KEY"

import requests

BASE_URL = "https://sandbox-api.coris.ai/api/v0_2"
HEADERS = {"Authorization": "Token YOUR_API_KEY"}

const BASE_URL = "https://sandbox-api.coris.ai/api/v0_2";
const HEADERS = {
  Authorization: "Token YOUR_API_KEY",
  "Content-Type": "application/json",
};

Fetch a page with totals. With metadata, the body is an object, not a plain array.

curl -s "$CORIS_BASE_URL/accounts/?limit=50&return_pagination_metadata=true" \
  -H "Authorization: Token $CORIS_API_KEY" | jq '.pagination_metadata.total'

page = requests.get(
    f"{BASE_URL}/accounts/",
    params={"limit": 50, "return_pagination_metadata": "true"},
    headers=HEADERS,
).json()
print(page["pagination_metadata"]["total"], len(page["data"]))

const pageResponse = await fetch(
  `${BASE_URL}/accounts/?limit=50&return_pagination_metadata=true`,
  { headers: HEADERS },
);
const page = await pageResponse.json();
console.log(page.pagination_metadata.total, page.data.length);

Try GET /accounts/ in the API explorer.

{
  "pagination_metadata": { "total": 142 },
  "data": [
    { "id": "c47f0e21-6a9b-4d38-a2e5-1b8d3f90c6a7", "reference_id": "acme-plumbing-001" }
  ]
}

pagination_metadata.total counts the items matching your filters, and can change during writes.

Walk the list with cursors, which suit large exports. Send sort_by=created&sort_order=desc on every request, and use the last id on each page as after_id.

AFTER=""
while : ; do
  PAGE=$(curl -s "$CORIS_BASE_URL/accounts/?limit=50&sort_by=created&sort_order=desc${AFTER:+&after_id=$AFTER}" \
    -H "Authorization: Token $CORIS_API_KEY")
  [ "$(printf '%s' "$PAGE" | jq 'length')" = "0" ] && break
  printf '%s' "$PAGE" | jq -r '.[].id'
  AFTER=$(printf '%s' "$PAGE" | jq -r '.[-1].id')
done

accounts, after_id = [], None

while True:
    params = {"limit": 50, "sort_by": "created", "sort_order": "desc"}
    if after_id:
        params["after_id"] = after_id
    rows = requests.get(f"{BASE_URL}/accounts/", params=params, headers=HEADERS).json()
    if not rows:
        break
    accounts.extend(rows)
    after_id = rows[-1]["id"]

print(len(accounts))

const accounts = [];
let afterId = null;

while (true) {
  const query = new URLSearchParams({ limit: "50", sort_by: "created", sort_order: "desc" });
  if (afterId) query.set("after_id", afterId);
  const rows = await (await fetch(`${BASE_URL}/accounts/?${query}`, { headers: HEADERS })).json();
  if (rows.length === 0) break;
  accounts.push(...rows);
  afterId = rows[rows.length - 1].id;
}
console.log(accounts.length);

Cursor rules:

A first page with no sort_by has no set order, so rows repeat or go missing.
before_id, after_id and offset cannot be combined.
An invalid cursor returns 400.

Pagination parameters

Parameter	Meaning	Default
`limit`	Items per page.	10
`offset`	Items to skip.	0
`after`	Items created after this epoch time.	None
`before`	Items created before this epoch time.	None
`return_pagination_metadata`	Send any value, even `false`, to get the object above. Leave it out to get a plain array.	Off

These lists ignore limit and return every row: /account_tags/, /settings/custom_list/, /accounts/{account_id}/persons/, /accounts/{account_id}/websites/ and /accounts/{account_id}/metadata/.

Rate limits

Limit	Value	Applies to
Authenticated requests	1,000 per minute, per user	Endpoints using the standard throttle
Concurrent synchronous data requests	25 by default, per workspace	`GET` requests that fetch data live

POST requests and GET requests with request_id skip the concurrency limit, so use the async flow for bulk work.

What to do next

Situation	Action
`429` with `Retry-After`	Wait that many seconds, then retry.
`429` with no `Retry-After`	You hit the concurrency limit. Back off for a few seconds and reduce parallel calls.
Your workload needs more concurrency	Ask for a higher limit.

FAQ

What is the maximum `limit`?

Most lists enforce no maximum, but stay near 100 so that responses stay fast.

Do all lists support cursors?

No. Only the accounts list uses them. Other lists use offset.

Should I fetch pages in parallel?

No. Fetch them in order, which also keeps your results consistent.

Next: Track API versions