Skip to main content
GET

Overview

Get text chunks from a text-type namespace via the public API. Use this to list stored text/summary chunks for display, export, or RAG. Results are returned one page at a time. Each page returns up to 100 items. When more data exists, the response includes pagination.has_more and a pagination.next_token — pass that token on the next request to fetch the following page.
Only namespaces with type === "text" are supported. Vector-only namespaces are not supported.

Authentication

string
required
Your API key. API Gateway requires a valid API key for this route.

Path Parameters

string
required
Name of the text namespace (e.g. my-docs)

Query Parameters

integer
Maximum number of items to return in this page. Default: 100. Maximum: 100.
string
Opaque cursor from the previous response’s pagination.next_token. Omit on the first request.
next_token is tied to your API key user and namespace. Do not reuse a token from a different namespace or account.

Pagination

1

First request

GET .../fetch-text-data?limit=100 (or omit limit for the default of 100).
2

Check pagination

If pagination.has_more is true, more chunks exist in the namespace.
3

Next page

Call again with next_token from the previous response: GET .../fetch-text-data?limit=100&next_token=<token>
4

Stop

Repeat until pagination.has_more is false or next_token is null.
Each page is a separate API call and counts toward your API request usage and credits.

Response Fields

Success Response (200)

string
Always "success" for 200 responses
string
Human-readable summary for this page (e.g. “Fetched N text items from namespace ’…’”)
string
The requested namespace name
object
Aggregated counts and timestamps for items in this page only (not the full namespace total)
number
Number of items in items on this page
number
Count of non-summary chunks on this page
number
Count of summary chunks on this page
string | null
Earliest item timestamp on this page as ISO 8601 (e.g. "2025-12-19T18:18:57.370Z"). Computed by normalizing each item’s stored createdAt (ISO string or Unix s/ms) and converting the minimum back to ISO. null when no item has a parseable date.
string | null
Latest item timestamp on this page as ISO 8601. Same normalization as created_at_min. null when no item has a parseable date.
object
Map of source name → count on this page. Keys may be file names, upload sources, or namespace identifiers.
array
Text chunks for this page (length ≤ limit, max 100)
string
Item/chunk identifier
string
Text content of the chunk
object
Arbitrary metadata (e.g. source, heading, chunk_index, page_title)
string | null
Creation time as an ISO 8601 string (e.g. "2025-12-19T18:18:57.700Z")
boolean
true for summary chunks, false for regular text chunks
object
Cursor pagination metadata for this response
number
Page size used for this request (≤ 100)
boolean
true if additional pages exist after this one
string | null
Pass as query param next_token to fetch the next page. null when has_more is false.
number
Request processing time in seconds

Limits

Items per page

Up to 100 items per request. Use limit and next_token to walk the full namespace.

Credits & API requests

Each page is one API call and consumes credits toward your plan’s API request limit.

Export all pages (example)