API reference

Knowledge base API

What an assistant answers from. Add pasted text, a single URL, a whole site or an uploaded file, watch indexing finish, and run the bot's own retrieval to see exactly what comes back for a query.

Endpoints

8 operations, each with its parameters, schemas and an example request on its own page.

get/api/bots/{bot_id}/knowledgeList documents
post/api/bots/{bot_id}/knowledgeCreate document — Adds pasted text, a single URL, or a whole site to the knowledge base.
post/api/bots/{bot_id}/knowledge/searchSearch knowledge — Runs the bot's own retrieval and shows exactly what comes back.
post/api/bots/{bot_id}/knowledge/uploadUpload document — Accepts a file upload.
delete/api/bots/{bot_id}/knowledge/{document_id}Delete document
get/api/bots/{bot_id}/knowledge/{document_id}/downloadDownload document — A short-lived signed link to the file the customer originally uploaded.
post/api/bots/{bot_id}/knowledge/{document_id}/reindexReindex document
get/api/bots/{bot_id}/knowledge/{document_id}/statusDocument status — Polled after upload until the document reaches a terminal state.

Authentication

Every call here needs Authorization: Bearer with a key from the dashboard — Authentication covers keys and scopes.

Objects returned

The models these endpoints send back, with every field. Open one to read it.

PaginatedResponse_DocumentRead_6 fields

itemsarray of DocumentReadrequired
14 fields · DocumentRead
- created_atstring (date-time)required
- idstringrequired
- source_typestringrequired
- statusstringrequired
- titlestringrequired
- updated_atstring (date-time)required
- chunk_countinteger
  Default: 0
- embedding_modelstring | null
- error_messagestring | null
- mime_typestring | null
- size_bytesinteger
  Default: 0
- source_urlstring | null
- tagsarray of any
- token_countinteger
  Default: 0
pageintegerrequired
page_sizeintegerrequired
totalintegerrequired
has_moreboolean
Default: false
next_cursorstring | null

DocumentRead14 fields

created_atstring (date-time)required
idstringrequired
source_typestringrequired
statusstringrequired
titlestringrequired
updated_atstring (date-time)required
chunk_countinteger
Default: 0
embedding_modelstring | null
error_messagestring | null
mime_typestring | null
size_bytesinteger
Default: 0
source_urlstring | null
tagsarray of any
token_countinteger
Default: 0

SearchResponse3 fields

querystringrequired
resultsarray of SearchResultrequired
8 fields · SearchResult
- chunk_idstringrequired
- contentstringrequired
- document_idstringrequired
- document_titlestring | nullrequired
- scorenumberrequired
- dense_rankinteger | null
- lexical_rankinteger | null
- rerank_scorenumber | null
took_msintegerrequired

Message2 fields

messagestringrequired
okboolean
Default: true

DocumentDownload4 fields

expires_inintegerrequired
filenamestringrequired
urlstringrequired
mime_typestring | null

DocumentStatusRead4 fields

chunk_countintegerrequired
idstringrequired
statusstringrequired
error_messagestring | null

Something here wrong or missing? Tell us — the documentation and the API are maintained by the same person, so a correction is a fix rather than a ticket.