Documentation for VectorDB Collection Management API¶
POST /collections/create — create_collection¶
Description¶
Creates a new vector collection for an embedding model. Also initializes a query-optimized collection.
Request¶
{
"embedding_model_id": "model-001"
}
Response¶
{
"embedding_model": { "id": "model-001" },
"index_status": "empty",
"collection_size": 0,
"last_updated": "2024-05-20T00:00:00Z"
}
POST /collections/create-index — create_index¶
Description¶
Creates HNSW indexes on both the main and query collections for the specified model.
Request¶
{
"embedding_model_id": "model-001"
}
POST /collections/categories/create-index — create_categories_index¶
Description¶
Same as create-index, but for category collections.
POST /collections/delete — delete_collection¶
Request¶
{
"embedding_model_id": "model-001"
}
POST /collections/categories/delete — delete_categories_collection¶
Request¶
{
"embedding_model_id": "model-001"
}
GET /collections/list — list_collections¶
GET /collections/queries/list — list_query_collections¶
GET /collections/categories/list — list_category_collections¶
Description¶
Returns metadata of all collections for general, query, or category types.
Response¶
{
"collections": [
{
"embedding_model": { "id": "model-001" },
"index_status": "built",
"collection_size": 12000,
"last_updated": "2024-05-20T00:00:00Z"
}
]
}
GET /collections/get-info¶
GET /collections/categories/get-info¶
Request¶
{
"embedding_model_id": "model-001"
}
POST /collections/set-blue¶
POST /collections/categories/set-blue¶
Description¶
Promotes the specified collection to "blue" (primary) status.
Request¶
{
"embedding_model_id": "model-001"
}
GET /collections/get-blue-info¶
GET /collections/get-blue-query-info¶
GET /collections/categories/get-blue-info¶
Description¶
Returns state of the active ("blue") collection for model/query/category.
POST /collections/objects/insert¶
POST /collections/categories/objects/insert¶
Request¶
{
"embedding_model_id": "model-001",
"objects": [
{ "object_id": "obj-1", "vector": [0.1, 0.2], "metadata": {}, "payload": {} }
]
}
POST /collections/objects/upsert¶
POST /collections/categories/objects/upsert¶
Request¶
{
"embedding_model_id": "model-001",
"shrink_parts": true,
"objects": [
{ "object_id": "obj-1", "vector": [0.1, 0.2], "metadata": {}, "payload": {} }
]
}
POST /collections/objects/delete¶
POST /collections/categories/objects/delete¶
Request¶
{
"embedding_model_id": "model-001",
"object_ids": ["obj-1", "obj-2"]
}
POST /collections/objects/find-by-ids¶
POST /collections/categories/objects/find-by-ids¶
Request¶
{
"embedding_model_id": "model-001",
"object_ids": ["obj-1", "obj-2"]
}
Response¶
[
{
"object_id": "obj-1",
"vector": [0.1, 0.2],
"metadata": {}
}
]
POST /collections/objects/find-similar¶
POST /collections/categories/objects/find-similar¶
Request¶
{
"embedding_model_id": "model-001",
"query_vector": [0.1, 0.2],
"limit": 5,
"offset": 0,
"max_distance": 0.5
}
Response¶
[
{
"object_id": "obj-1",
"distance": 0.123,
"vector": [0.1, 0.2],
"metadata": {}
}
]