> For the complete documentation index, see [llms.txt](https://docs.tiny.plus/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.tiny.plus/api/concepts/fair-use.md).

# Fair use

Every tiny+ customer is served by the same system, so we ask every integration to use the API fairly. Most of the API isn't rate limited, so these guidelines are what keep it fast for everyone.

## Guidelines

* **Keep to about 60 requests a minute** for each access token, and **send no more than two requests at a time.**
* **Fetch changes, not everything.** To keep another system in step with tiny+, ask only for records changed since your last run, instead of reading every record each time. Filter on `modified_date`, for example `modified_date=>2026-09-30 00:00:00` (in UTC).
* **Poll no more often than every five minutes.**
* **Ask only for what you need.**
  * Pass `with_related=0` unless you need related records.
  * Use `fields=` to return only the fields you use.
  * Page through long lists a few hundred records at a time, rather than asking for the maximum of 1000.
* **Keep a copy of your account's structure.** Fields, statuses and tags rarely change. Read `/v2/fields`, `/v2/record_status` and `/v2/form_structure` once a day, not before every request.
* **Back off when tiny+ is busy.** If a request fails with a `429` or a `5xx` status, wait before trying again, and wait longer each time it fails.
* **Talk to us before a bulk job.** If you plan to create or update thousands of records, email <support@tiny.plus> first. An [import](/guides/getting-started/importing-your-content.md) may suit the job better.

If an integration's traffic slows tiny+ down for other customers, we may limit or suspend its access token. Where we can, we'll contact the account before we do.

## Limits we enforce

A few parts of the API have limits of their own:

| What                                      | Limit                                                                                                                                                                                                                                                                                                                                    |
| ----------------------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| Ask filters (`POST /v2/ai/filter`)        | These count against the account's [Ask allowance](/guides/everyday/ask-and-ai-assistants.md#fair-use), plus 30 AI requests a minute per person (counted per clock minute, and shared with AI report and automation design). Over either limit, the response is still `200`, with `"reason": "capped"` and a plain search to use instead. |
| AI assistants (`/mcp`)                    | 60 requests a minute for each person and assistant, 20 of them changes.                                                                                                                                                                                                                                                                  |
| Registering apps (`POST /oauth/register`) | 20 registrations an hour from the same address, with a `429` status beyond that.                                                                                                                                                                                                                                                         |