SnozeDevelopers
API referenceRag

List the embedding models registered for a workspace

Returns the embedding models this workspace has content indexed under, so a caller can tell which chunks live in comparable vector spaces. Requires a workspace member with the `ai:read` permission.

GET
/v1/workspaces/{workspaceId}/rag/models

Authorization

apiKey
headerAuthorizationBearer <snoze_sk_…>

Workspace API key created in Workspace settings → API keys. Scopes on the key bound what it can read and write.

Path Parameters

workspaceId*string

Response Body

Success

application/json
  1. response
data*array<>
Example Request
curl -X GET 'https://api.snoze.dev/v1/workspaces/string/rag/models'
Example Response
{  "data": [    {      "id": "string",      "slug": "string",      "label": "string",      "dimension": 0,      "isDefault": true,      "createdAt": null    }  ]}

Ingest a source document into the retrieval index

Chunks and embeds the supplied text, replacing any chunks previously indexed for the same `sourceKind` and `sourceId`, and returns the chunk count and the embedding model used. Requires a workspace member with the `ai:write` permission. Embedding spend is metered against the workspace AI allowance, but metering is deliberately fail-open and never blocks the ingest.

List the retrieval sources a chat can draw on

Returns every indexed source in the workspace with its label and chunk count, each flagged with whether this chat currently has it selected. Requires a workspace member with the `ai:read` permission; a chat that does not exist in this workspace is reported as not found.