/store-and-search-documents

Store Document

Upload a document to have it stored and searchable.
You can provide a file, a public URL, or raw text.
You can also attach optional metadata filters.

Request Body

Send data as multipart/form-data.

Accepted File Types

You can attach up to 7 distinct filter fields per document.
These must be formatted as a flat JSON object in the filters field.
You must use these exact keys:

• number_1, number_2, number_3
• string_1, string_2, string_3, string_4

Example:

{
  "string_1": "hr-policies",
  "string_2": "internal",
  "string_3": "onboarding",
  "string_4": "v2",
  "number_1": 2025,
  "number_2": -3.14159,
  "number_3": 42
}

Document Fields

Every stored document contains the following fields.

Search Documents

Perform a semantic vector search across all documents in a dataspace.
You can also apply complex metadata filters to narrow down the results before ranking.

Request Body

Send data as application/json.

Filter Rules

Each rule in the filters array must be an object with three properties: field, operator, and value.
Operators can be combined freely.

Limits:

• · You can use maximum 7 filter rules in a request.
• · You can only use one in or not-in operator per search query.
• · You can only use one != or not-in filter per search query.

Request Examples

Basic query

{
  "query": "What is the remote work policy?"
}

With limit

{
  "query": "What is the remote work policy?",
  "limit": 5
}

Single filter

{
  "query": "How do I cancel my subscription?",
  "limit": 5,
  "filters": [
    { "field": "string_1", "operator": "==", "value": "support" }
  ]
}

Multiple filters

{
  "query": "GDPR compliance requirements for user data",
  "limit": 10,
  "filters": [
    { "field": "string_1", "operator": "==", "value": "legal" },
    { "field": "number_1", "operator": ">=", "value": 2024 },
    { "field": "string_2", "operator": "!=", "value": "draft" }
  ]
}

Range + in operator

{
  "query": "pricing model",
  "limit": 20,
  "filters": [
    { "field": "number_2", "operator": ">=", "value": 3.14 },
    { "field": "number_2", "operator": "<", "value": 3.15 },
    { "field": "string_4", "operator": "in", "value": ["finance", "sales", "marketing"] }
  ]
}

Other Endpoints

Detailed specifications for all remaining CRUD and management endpoints.

Pagination

Both GET /dataspaces and GET /dataspaces/:id/documents support cursor-based pagination.

Every paginated response includes a pagination object:

{
  "status": "success",
  "data": [ ... ],
  "pagination": {
    "has_more": true,
    "next_cursor": "8f3kLmNpQ2xR4vW1"
  }
}

List Documents

Returns a paginated list of documents in a dataspace, ordered by creation date (newest first).

{
  "status": "success",
  "data": [
    {
      "id": "8f3kLmNpQ2xR4vW1",
      "markdown": "# Company Wiki\n\n## Remote Work\nEmployees are allowed...",
      "text": "Company Wiki\nRemote Work\nEmployees are allowed...",
      "filters": {
        "string_1": "hr-policies",
        "number_1": 2025
      },
      "created_at": "2026-04-30T12:00:00Z"
    },
    {
      "id": "Kp9nWxYz5TmR3qL7",
      "markdown": "",
      "text": "A monthly stipend of $200 is provided for home office equipment...",
      "filters": {
        "string_1": "hr-policies",
        "number_1": 2025
      },
      "created_at": "2026-04-30T11:45:00Z"
    }
  ],
  "pagination": {
    "has_more": true,
    "next_cursor": "Kp9nWxYz5TmR3qL7"
  }
}

Get Document

Retrieve a single document by its ID.

{
  "status": "success",
  "data": {
    "id": "8f3kLmNpQ2xR4vW1",
    "markdown": "# Company Wiki\n\n## Remote Work\nEmployees are allowed...",
    "text": "Company Wiki\nRemote Work\nEmployees are allowed...",
    "filters": {
      "string_1": "hr-policies",
      "number_1": 2025
    },
    "created_at": "2026-04-30T12:00:00Z"
  }
}

Update Document

Update metadata filters only, on an existing document.
Send data as application/json with a filters object.

• · Partial merge: Only the keys you include are updated. Existing filters you don't mention are left unchanged.
• · Delete a filter: Send null for a key to remove it entirely.
• · Content is immutable: You cannot update the text, file, or markdown of a document. Delete and re-store it instead.

Request body

{
  "filters": {
    "string_1": "updated-category",
    "number_1": 2026,
    "string_2": null
  }
}

Response

{
  "status": "success",
  "data": {
    "id": "8f3kLmNpQ2xR4vW1",
    "markdown": "# Company Wiki\n\n## Remote Work\nEmployees are allowed...",
    "text": "Company Wiki\nRemote Work\nEmployees are allowed...",
    "filters": {
      "string_1": "updated-category",
      "number_1": 2026
    },
    "created_at": "2026-04-30T12:00:00Z"
  }
}

Note that string_2 was removed from the response because it was set to null in the request. Filter updates take effect on the next search immediately.

Delete Document

Permanently delete a document and all its associated search data.

{
  "status": "success",
  "data": {
    "id": "8f3kLmNpQ2xR4vW1",
    "deleted": true
  }
}

Create Dataspace

Create a new, empty dataspace. No request body is required.

{
  "status": "success",
  "data": {
    "id": "Xr9pLmWq4TnK2vY8",
    "count": 0,
    "created_at": "2026-04-30T14:00:00Z"
  }
}

List Dataspaces

Returns a paginated list of your dataspaces, ordered by creation date (newest first).
Supports the same ?limit and ?cursor query parameters.

{
  "status": "success",
  "data": [
    {
      "id": "Xr9pLmWq4TnK2vY8",
      "count": 47,
      "created_at": "2026-04-30T14:00:00Z"
    },
    {
      "id": "Bc3nYhTw7KpR5jM1",
      "count": 12,
      "created_at": "2026-04-28T09:30:00Z"
    }
  ],
  "pagination": {
    "has_more": false,
    "next_cursor": "Bc3nYhTw7KpR5jM1"
  }
}

Get Dataspace

Retrieve a single dataspace. The count field reflects the number of documents currently stored.

{
  "status": "success",
  "data": {
    "id": "Xr9pLmWq4TnK2vY8",
    "count": 47,
    "created_at": "2026-04-30T14:00:00Z"
  }
}

Delete Dataspace

Permanently delete a dataspace. All documents inside it are permanently destroyed.

{
  "status": "success",
  "data": {
    "id": "Xr9pLmWq4TnK2vY8",
    "deleted": true,
    "documents_destroyed": 47
  }
}

Limits

Document storage limit: The total number of documents you can store across all dataspaces depends on your plan — 200 on the Free plan, 50,000 on Basic, 100,000 on Plus, and 200,000 on Max.