Search Seeds

Perform semantic search across seeds using natural language queries with vector embeddings.

Endpoint

POST /seeds/query

Required. Include your API key in the Authorization header:

Authorization: Bearer YOUR_API_KEY

Parameter	Type	Required	Description
appId	string	Yes	Application identifier
externalUserId	string	Yes	External user identifier for data isolation

Field	Type	Required	Description
query	string	Yes	Natural language search query
seedIds	string[]	No	Limit search to specific seed IDs (default: [])
limit	number	No	Max results to return (1-100, default: 30)
threshold	number	No	Similarity threshold (0-1, default: 0.5)
imageData	string	No	Base64-encoded image for image search (data:image/...)

Field	Type	Description
results	array	Array of search results
results[].seedId	string	ID of the seed containing the match
results[].chunkId	string	ID of the matching chunk
results[].content	string	Content of the matching chunk
results[].similarity	number	Similarity score (0-1, higher is better)
total	number	Total matching results
queryTime	number	Query execution time in seconds

Code	Description
200	Success - Search results returned
400	Bad Request - Invalid query or parameters
401	Unauthorized - Invalid API key
500	Internal Server Error

Jina Embeddings v4: Multimodal embeddings supporting both text and image search in unified vector space
Redis caching: Hot chunks cached for 10-50x faster subsequent queries
Seed filtering: Restrict search to specific seeds using seedIds parameter
Similarity threshold: Control result quality with threshold parameter
Image search: Search using images as input with imageData parameter

Base URL: All requests use https://api-development.myneutron.ai which adapts to your environment.