List Available Models

GET

/api/discovery/enhanced-search/models

• DivinityEngine - Local development
• DivinityEngine - Localhost alternative

List the embedding models the bundled ONNX runtime can serve.

Returns:

• available_models: list of embedding model names the runtime exposes
• current_model: currently active embedding model name

Bundled today:

• multilingual-e5-large-instruct (1024 dimensions) — single bundled ONNX model. Older identifiers like all-mpnet-base-v2 are accepted for backward compatibility but resolve to the same E5 weights at runtime.

Responses

200

Successful Response

SearchModelsResponse

Search models listing response model.

Provides listing of embedding models the bundled ONNX runtime can serve for enhanced semantic search operations.

Fields:

• available_models: List of embedding model names the runtime exposes
• current_model: Currently active embedding model name

Bundled today:

• multilingual-e5-large-instruct (1024 dimensions) — the single bundled ONNX model. Older identifiers like all-mpnet-base-v2 and all-MiniLM-L6-v2 resolve to the same weights at runtime for backward compatibility.

Usage: GET /api/discovery/enhanced-search/models returns this response model.

JSON Example:

{
  "availableModels": ["multilingual-e5-large-instruct"],
  "currentModel": "multilingual-e5-large-instruct"
}

object

availableModels

Availablemodels

Available embedding model names

Array<string>

currentModel

required

Currentmodel

Currently active embedding model

string