Skip to main content
Base Node Node inspector for a MongoDB Vector Search node

Parameters

Hostname Scheme
string
default:"mongodb+srv"
required
Scheme to use when constructing the MongoDB connection URI. Use mongodb or mongodb+srv.
Host
string
default:"localhost"
required
MongoDB host or SRV cluster hostname.
Port
integer
MongoDB port to connect to. Leave empty to use the default port for the selected scheme.
App Name
string
Optional appName query parameter for the MongoDB connection URI.
Database
string
default:"default"
required
Database name containing the vector collection.
Collection
string
default:"documents"
required
Collection name containing the vector index.
Index Name
string
default:"vector_index"
required
Name of the vector search index.
Vector Field
string
default:"embedding"
required
Field containing the stored embeddings.
Number of Candidates
integer
default:"100"
required
Number of candidates to consider during vector search. Must be at least the Limit and no more than 10000. It is recommended to set this at least 20x higher than the Limit for best results.
Limit
integer
default:"5"
required
Number of documents to return. Must be less than or equal to the Number of Candidates.
Embedding Model
string
default:"text-embedding-3-small"
Embedding model used to encode queries. The model name should be in the format expected by LiteLLM. Additionally, it should match the model used to index the vectors in the MongoDB collection.
Include Fields
array
default:"[]"
Optional list of fields to include in results. Leave empty to return full documents.
Include Score
boolean
default:"True"
Include vector search scores in the output.
Tool Name
string
default:"mongodb_vector_search"
required
Name to register for the generated tool.
Tool Description
string
required
Description to register for the generated tool.

Credentials

Password Credential
Password Credential
required
Select a Password Credential containing the MongoDB username and password.
LLM Credential
LLM Credential
required
Select an LLM Credential used to generate query embeddings.

Inputs

This node has no inputs.

Outputs

Output
tool
Generated tool configuration for the Agent node.