Search that understands meaning.
Runs in the browser. Completely private.

Offline semantic search for any documentation site. Understands meaning, runs entirely in the browser, and never sends data to the cloud.

Get Started

Local Vector Engine 100% Private Quantised Index Zero Cloud Costs

user authentication

⌘K

User Authentication & Sessions

/getting-started/security

98% match

Configure how your application handles user login, secure cookies, token generation, and stateless session verification.

OAuth & Single Sign-On (SSO)

/plugins/auth-providers

92% match

Authenticate documentation access through GitHub, Google, and enterprise OAuth SAML providers.

Route Guards & Private Docs

/configuration/routing

86% match

Define private directories and redirect configurations for unauthenticated guest sessions trying to access protected paths.

Architecture

How it works

Smart search for static documentation, with no servers or third-party services required.

Runs Locally in the Browser

Uses lightweight AI models that execute directly in the browser. No data ever leaves your users' devices.

Int8 Quantised Vectors · Zero Telemetry · <3 KB Client

Understands What You Mean

Goes beyond exact keyword matching. Searching for 'authentication' also finds results about 'sign-in', 'login', and 'sessions'.

Dense 384-Dim Embeddings · Cosine Similarity Re-ranker

Hybrid Keyword + Semantic

Combines fast keyword matching with meaning-aware similarity scoring for the most relevant results, instantly.

BM25 Lexical + Cosine Vector Fusion · Typo-Tolerant

Comparison

How docmd-search compares

Client-side AI vector search vs traditional cloud APIs and static keyword engines.

Capability	docmd Search	Algolia / Cloud Search	Pagefind / Lunr
Semantic Understanding Finds concepts by meaning, not just exact keyword strings	Dense Vector AI	NeuralSearch (Enterprise)	Keyword / Lexical Only
Infrastructure & Monthly Bill Ongoing hosting, vector DB, or API query subscription costs	$0 / Forever Free	Usage-Based / $$$ Contract	$0 / Forever Free
Offline & Air-Gapped Operation Runs on intranets, offline PWA, local dev without internet	100% Offline	Requires Internet	100% Offline
User Privacy & Data Residency Where queries are processed and logged	Zero Data Sent	Queries Logged to Cloud	Zero Data Sent
Search Latency Response speed per typed keystroke	< 1ms (In-Memory)	50–250ms (Network)	< 5ms (In-Memory)

Embeddings

Choose your search model

Pick the model that fits your documentation size and language requirements.

Model	Dimensions	Size	Languages	Best For
MiniLM L6 v2 Recommended ★	384	~23 MB	English only	Fast, general English docs
Multilingual MiniLM L12	384	~118 MB	50+ languages	i18n documentation
Multilingual E5 Small	384	~118 MB	100+ languages	Wide language coverage
Multilingual MPNet Base	768	~270 MB	50+ languages	Best multilingual quality

💡 Multilingual documentation: If your documentation website contains multiple languages (such as English, Chinese, German, Spanish, etc.), select a multilingual model using docmd-search --settings. The default model is English-only and will produce poor search relevance for other languages.

Developer API

Use it anywhere: the client API

A lightweight client runtime (<3 KB gzipped) that runs entirely in the browser. Build custom search UIs for any website or application.

Under 3 KB Gzipped

Pure lightweight JavaScript. No WASM, no model downloads, zero third-party dependencies.

Progressive Batch Streaming

Batch 0 loads instantly for immediate search while remaining chunks stream in background.

Framework Agnostic

Drop into React, Vue, Next.js, Astro, or static HTML. Full TypeScript definitions included.

100% Local Privacy

Evaluates queries in device memory with zero telemetry. Fully compliant with strict GDPR policies.

search-client.ts

import { load, search } from 'docmd-search/client';

// 1. Initialise index (batch 000 loads instantly, rest stream in background)
await load('/_docmd-search', (loaded, total) => {
  console.log(`Loaded batch ${loaded}/${total}`);
});

// 2. Query search with hybrid vector + keyword scoring (<1ms)
const results = search('authentication secure routes', 5);
results.forEach(({ score, chunk }) => {
  console.log(`[${Math.round(score * 100)}%] ${chunk.file}#${chunk.heading || ''}`);
});

Console Output

[0.3ms] [98%] /getting-started/security#authentication

[0.4ms] [92%] /plugins/auth-providers#oauth-sso

Using docmd? Enable semantic search directly in docmd.config.json under "plugins": { "search": { "semantic": true } } – no custom JavaScript required.

Frequently asked questions

Common questions about docmd-search and offline semantic search.

Does it run entirely in the browser?

Yes. All search happens directly in your users' browsers. No cloud infrastructure, no search API, and no data ever leaves the device.

Read the full documentation →

Do users need to download AI models?

No. The AI models are only used at build time on your machine to generate pre-computed search indices. Users' browsers only download the resulting index files – small, compressed JSON chunks.

Learn how offline search works →

Does it work with multiple languages?

Yes. For multilingual documentation, choose a multilingual model such as paraphrase-multilingual-MiniLM-L12-v2, which supports over 50 languages. The default model is English-only.

Configure multilingual search →

What is the confidence score badge?

When enabled, the showConfidence setting displays a percentage badge next to each search result showing how closely it matches the query. This helps users quickly judge which results are most relevant.

Read about browser client settings →