Vector DB
Search
Run k-nearest-neighbor similarity search against a collection.
Search by vector
curl -X POST https://api.openadapter.in/v1/vectors/collections/docs/search \
-H "Authorization: Bearer $OPENADAPTER_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"vector": [0.01, -0.02, ...],
"limit": 5
}'Response:
{
"results": [
{
"id": 7,
"score": 0.91,
"payload": { "text": "..." }
}
]
}| Field | Notes |
|---|---|
vector | Length must equal the collection's vector_size |
limit | Number of nearest points to return |
Notes
scoreis the similarity from whatever distance metric the collection was created with.- Each vector API call costs 0.05 of one request against your plan quota.
- Payload filtering and filter-based deletes may be supported by the upstream vector service. The gateway is a passthrough, so the request body is forwarded as-is — but the shape of any filter object is not currently documented by the dashboard. Check
GET /v1/vectors/collections/{name}for the live collection config before relying on filter features.