Skip to main content
Search returns assets from an index. Results depend on the features you ran during indexing (for example embeddings and transcriptions). The current API exposes a placeholder /search/ endpoint and may evolve; use it to experiment and prefer filtering assets for now.

​
Experimental: POST /search/

curl -X POST "$ASPECT_URL/search/" \
  -H "Authorization: Bearer $ASPECT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{}'
Response example: 
{
  "results": [
    { "id": "result_1", "type": "asset", "title": "Sample Video", "relevance_score": 0.95 },
    { "id": "result_2", "type": "group", "title": "Sample Group", "relevance_score": 0.87 }
  ],
  "total_count": 2,
  "query_time_ms": 150
}
Note: This endpoint is currently a stub. For production flows today, fetch assets by index and use your application-side filters or metadata.

​
List assets in an index

curl -X GET "$ASPECT_URL/assets?index_id=<INDEX_UUID>" \
  -H "Authorization: Bearer $ASPECT_API_KEY"
Each asset includes features states so you can determine whether embeddings, transcriptions, and proxies are ready for your own filtering or retrieval logic.
Track the API reference for updates to the search API, including request parameters and scoring models.