AI-powered search

Hybrid Search

Match how people actually search. Combine keyword precision with AI that understands meaning and intent.

Trusted by leading companies

Perfect for

Anywhere users search with both exact terms and natural language.

A custom shopping experience for every buyer

Every customer gets results tailored to their preferences. After just a few searches or clicks, shoppers see products that match their style, budget, and interests.

↑ 23%Conversion rate
↓ 40%Zero-result searches
↑ 15%Average order value

Search that understands intent

When users know exactly what they want, we find it. When they're exploring, we help them discover. Meilisearch automatically detects intent and adapts.

QueryIntentKeywordSemanticHybrid
"Nike Air Max 90"Known item
Exact match found
May miss exact
KeywordKeyword-firstSemantic
"comfortable office footwear"Discovery
Few or no results
Finds related
KeywordSemantic-firstSemantic
"running shoes"Ambiguous
Exact matches only
Related concepts
KeywordBalancedSemantic

The best of both worlds

Keyword precision meets semantic understanding, tuned to your needs.

Understands meaning

"Affordable housing" matches "budget apartments". Find related results even when wording differs.

Tunable mix

Dial up keyword precision or semantic understanding for each use case. One setting, no code changes.

Any AI provider

Native integrations with OpenAI, Mistral, and more. Bring your own model through a simple endpoint.

Zero setup

Embeddings happen automatically when you add a document. No extra pipeline to build or maintain.

Scales to millions

Compressed vectors keep search fast and storage costs low, even with millions of documents.

Control what AI sees

Choose exactly which fields feed your AI model. Better inputs mean more accurate matches.

Works with all major embedding providers

Native integrations, no backend required for most providers.

38 models from 11 providers

OpenAI

OpenAI

2 models

text-embedding-3-large
text-embedding-3-small
Google

Google

3 models

gemini-embedding-2
gemini-embedding-001
text-embedding-005
Mistral

Mistral AI

1 model

Mistral Embed
Cohere

Cohere

3 models

Embed 4
Embed v3
Embed Multilingual v3
HuggingFace

Hugging Face

5 models

sentence-transformers
all-MiniLM-L6-v2
all-mpnet-base-v2
e5-large-v2
multilingual-e5-large
Ollama

Ollama

5 models

nomic-embed-text
mxbai-embed-large
bge-m3
all-minilm
snowflake-arctic-embed
together.ai

Together AI

1 model

BAAI bge-large
Cloudflare

Cloudflare AI

1 model

bge-m3
Voyage

Voyage AI

10 models

voyage-4-large
voyage-4
voyage-4-lite
voyage-4-nano
voyage-3.5
voyage-3.5-lite
voyage-multimodal-3.5
voyage-code-3
voyage-finance-2
voyage-law-2
Jina

Jina AI

4 models

jina-embeddings-v4
jina-embeddings-v3
jina-clip-v2
jina-colbert-v2
Mixedbread AI

Mixedbread AI

3 models

mxbai-embed-large-v1
mxbai-embed-2d-large-v1
mxbai-colbert-large-v1
+

Custom

Any provider

Meilisearch is compatible with any model offering a REST API and tool calling capabilities.

Frequently asked questions

Hybrid search combines traditional keyword search with AI-powered semantic search. Keywords find exact matches, while semantic search understands meaning and intent. Together, they deliver more relevant results than either approach alone.

Ready to get started?

Run Hybrid Search on Meilisearch Cloud, or self-host the open-source engine.