Everything you need for production search

LumoSearch combines BM25F relevance ranking, typo tolerance, and instant autocomplete in a single, easy-to-deploy search engine.

BM25F Ranking

Industry-standard ranking algorithm that scores documents based on term frequency, inverse document frequency, and field-level boosting. Delivers highly relevant results out of the box without manual tuning.

  • Per-field length normalization
  • Configurable k1 and b parameters
  • Automatic IDF calculation across your corpus

Fuzzy Matching

Handles typos and misspellings gracefully using trigram-based similarity. Users find what they need even when they can't spell it correctly.

  • Trigram Jaccard similarity scoring
  • Configurable fuzziness threshold
  • Works alongside exact and prefix matching

Instant Autocomplete

Sub-50ms prefix-based autocomplete for real-time search-as-you-type interfaces. Built on a dedicated prefix index for maximum speed.

  • Prefix index for instant lookups
  • Results ranked by relevance, not just alphabetical
  • Configurable result limits

Simple REST API

A developer-first REST API that feels natural in any tech stack. Two endpoints: search and autocomplete. API keys for authentication. JSON in, JSON out.

  • POST /v1/search and POST /v1/autocomplete
  • Admin keys for management, public keys for search
  • Official SDK for JavaScript/TypeScript

Weighted Fields

Assign custom weights to different fields so titles matter more than body text, or tags rank higher than descriptions. Fine-tune relevance without changing your data.

  • Per-field weight configuration at collection creation
  • Weights applied during BM25F scoring
  • Combine with highlight ranges for precise UX

Highlight Ranges

Automatically identifies and returns the exact character positions of matching terms within your documents. Build search result snippets with precise highlighting.

  • Character-level offset ranges
  • Works across all matching strategies
  • Highlight any searchable field

Secure by Default

API keys are hashed with SHA-256 before storage. Admin keys for management, scoped public keys for search. Keys are shown once at creation and never stored in plain text.

  • SHA-256 hashed key storage
  • Collection-scoped public search keys
  • Prefix-based key lookup (no full scan)

Hosted Infrastructure

Deploy on LumoSearch Cloud and never think about servers. Automatic scaling, EU-hosted data, encrypted at rest and in transit. Focus on your product.

  • AWS infrastructure (eu-central-1)
  • Auto-scaling Lambda-based architecture
  • 5-minute warm cache for low latency

Usage Analytics

Track search volume, document counts, and API usage in real-time. Understand how your users search and when you need to scale.

  • Daily usage counters
  • Per-account search and upload tracking
  • Dashboard with usage visualization

Ready to add search to your app?

Get started with the open-source engine or join the waitlist for LumoSearch Cloud.

View on GitHub