Skip to main content
Some APIs return more results than fit in one response, such as an account’s followers. Pagination support is SKU-specific.

Confirm pagination support

Fetch the selected API’s contract with GET /v1/apis/{sku} or MCP get_api. Use cursor pagination only when inputSchema declares cursor and the output schema declares nextCursor. Include limit or requireSinglePage only when inputSchema also declares those fields. Call again with data.nextCursor as cursor until nextCursor is null. The examples use instagram.followers, whose schema declares these fields.

What the fields mean

  • limit, when declared, is the most items you want in this response: a page cap, not a grand total. Follow nextCursor for more.
  • cursor is an opaque token from a previous response’s nextCursor. Pass it to continue, or omit it to start from the first page. Do not parse or build it: it is sealed and tied to one walk.
  • nextCursor is in the response at output.data.nextCursor. A non-null value means there is more; null means you have reached the end of this walk.
  • page, on APIs whose schema declares it (today google.search), picks a page directly: send page: 2, page: 3 and so on, each billed as one request. Any source can serve any page. On those APIs cursor is deprecated but still accepted.
The selected SKU’s schema defines its page-size constraints. Do not copy a limit value from another API. If a cursor is rejected because it expired, start again without a cursor.

Get everything in one call

If you cannot loop and the selected inputSchema declares requireSinglePage, set requireSinglePage: true to request up to the schema’s allowed limit in one response. This route can cost more than walking pages.
curl

Cost

Check the selected API’s pricing before choosing between paging and a supported single-page request. Your exact charge in USD is returned as costUsd.