Meilisearch Has a Free API: Lightning-Fast Search Engine for Your Apps

#search #api #opensource #tutorial

Meilisearch is an open-source, lightning-fast search engine that provides instant search-as-you-type experiences. With typo tolerance, faceted search, and a RESTful API, it integrates into any application in minutes.

What Is Meilisearch?

Meilisearch is a Rust-based search engine designed for end-user facing search. It returns results in under 50ms, handles typos, and supports multiple languages out of the box. Think Algolia, but open-source and self-hosted.

Key Features:

Search in under 50ms
Typo tolerance
Faceted search and filtering
Geo search
Multi-language support
Synonyms and stop words
Tenant tokens (multi-tenancy)
RESTful API

Quick Start

# Install and run
curl -L https://install.meilisearch.com | sh
./meilisearch --master-key="your-master-key"

# Or via Docker
docker run -d -p 7700:7700 -v meili_data:/meili_data \
  getmeili/meilisearch:latest \
  meilisearch --master-key="your-master-key"

Meilisearch API: Index and Search

import requests

MEILI = "http://localhost:7700"
HEADERS = {"Authorization": "Bearer your-master-key", "Content-Type": "application/json"}

# Create index and add documents
documents = [
    {"id": 1, "title": "React Tutorial", "category": "frontend", "views": 15000},
    {"id": 2, "title": "Docker for Beginners", "category": "devops", "views": 23000},
    {"id": 3, "title": "Python Web Scraping Guide", "category": "backend", "views": 18000},
    {"id": 4, "title": "Kubernetes Best Practices", "category": "devops", "views": 12000},
    {"id": 5, "title": "TypeScript Advanced Types", "category": "frontend", "views": 9000}
]

requests.post(f"{MEILI}/indexes/articles/documents", headers=HEADERS, json=documents)

# Search (with typo tolerance!)
result = requests.post(f"{MEILI}/indexes/articles/search", headers=HEADERS, json={
    "q": "kuberntes",  # typo in 'kubernetes' - still finds it!
    "limit": 10
}).json()
for hit in result["hits"]:
    print(f"{hit['title']} ({hit['category']}) - {hit['views']} views")

# Faceted search
result = requests.post(f"{MEILI}/indexes/articles/search", headers=HEADERS, json={
    "q": "",
    "facets": ["category"],
    "filter": "category = devops AND views > 10000",
    "sort": ["views:desc"]
}).json()
print(f"Facets: {result['facetDistribution']}")

Settings Configuration

# Configure searchable and filterable attributes
requests.patch(f"{MEILI}/indexes/articles/settings", headers=HEADERS, json={
    "searchableAttributes": ["title", "category"],
    "filterableAttributes": ["category", "views"],
    "sortableAttributes": ["views"],
    "synonyms": {
        "k8s": ["kubernetes"],
        "js": ["javascript"],
        "ts": ["typescript"]
    },
    "stopWords": ["the", "a", "an", "is", "for"]
})

Multi-Search

# Search multiple indexes at once
results = requests.post(f"{MEILI}/multi-search", headers=HEADERS, json={
    "queries": [
        {"indexUid": "articles", "q": "docker", "limit": 3},
        {"indexUid": "tutorials", "q": "docker", "limit": 3}
    ]
}).json()

Python SDK

import meilisearch

client = meilisearch.Client("http://localhost:7700", "your-master-key")
index = client.index("articles")

# Add documents
index.add_documents(documents)

# Search
results = index.search("react tutorial", {"limit": 5, "attributesToHighlight": ["title"]})
for hit in results["hits"]:
    print(hit["_formatted"]["title"])  # Highlighted matches

Resources

Need to scrape web data for your search engine? Check out my web scraping tools on Apify — production-ready actors for Reddit, Google Maps, and more. Questions? Email me at spinov001@gmail.com