Hybrid search combining vector similarity and PostgreSQL full-text search, merged with Reciprocal Rank Fusion (RRF). Query variants are generated automatically to improve recall.
API key created from the MemContext dashboard. Keys are prefixed with mc_.
Natural language search query. Use full sentences for best results.
1000Maximum number of results to return.
1 <= x <= 10Filter results to a single category.
preference - User likes or dislikes (e.g. "prefers dark mode").fact - Objective information (e.g. "uses macOS").decision - Choices made (e.g. "chose PostgreSQL for DB").context - Background information (e.g. "working on e-commerce app").preference, fact, decision, context Filter results to a specific project.
Maximum vector distance. Higher values return more results (broader search).
Use 0.2-0.4 for strict matching, 0.6 as default, 0.7-0.8 for broad recall.
0 <= x <= 1