Access Your Configs
The configuration for HelixDB is a json file named config.hx.json.
This file will be created when you run helix init --path helixdb-cfg.
Config Options
Vector Configs
m: Maximum number of connections each node can have at each layer of the graph (default 16)ef_construction: Size of dynamic list of candidates considered during vector insertion (default 128)ef_search: Size of the dynamic list of candidates maintained during the search process (default 768)
| Parameter | Effect | Higher Value | Lower Value |
|---|
m | Controls graph connectivity | Richly connected graph with better pathfinding but more memory | Sparse graph with faster construction but worse pathfinding |
ef_construction | Size of candidate list during indexing | High-quality, accurate index but more indexing time and memory | Faster indexing but can miss some connections & degrade recall |
ef_search | Candidate pool size during query | Wider search scope, improving recall and accuracy but slower search | Narrower search scope, faster search but more approximate results |
Graph Configs
secondary_indices: List of secondary indices for graph operations (default empty)
Database Configs
db_max_size_gb: The maximum size of the database in GB (default 10)mcp: Enable MCP support (default true)embedding_model: Define what embedding model to use when calling Embed in the query languagegraphvis_node_label: Define what property label to display on nodes in the graph visualizer (run with helix visualize <INSTANCE_ID>)
All vectors in a vector type must have the same dimensions. If you change your embedding model, the new vectors will have different dimensions and will cause an error. Ensure you use the same embedding model consistently for all vectors.
