Search capabilities
Full-text search
Lightning-fast keyword search with typo tolerance and customizable relevancy.| Feature | Description |
|---|---|
| Sub-50ms responses | Fast search regardless of dataset size |
| Typo tolerance | Get relevant results even with spelling mistakes |
| Ranking rules | 6 default rules plus custom ranking |
| Ranking score | Relevancy scores with optional detailed breakdown |
| Synonyms | Define equivalent terms for better recall |
| Stop words | Ignore common words like “the” or “and” |
| Distinct attribute | Deduplicate results by a specific field |
| Prefix search | Results update as users type |
| Matching strategy | Control how query terms are matched: last, all, or frequency |
| Phrase search | Use double quotes to search for an exact phrase |
| Negative search | Exclude terms from results using the minus operator |
| Placeholder search | Return all results when the query is empty |
Querying and result formatting
| Feature | Description |
|---|---|
| Highlighting | Emphasize query matches in results with customizable tags |
| Cropping | Return only the relevant portion of long text fields |
| Matches position | Get byte positions and lengths of matched terms |
| Search cutoff | Set a maximum time limit for search queries |
| Tokenization | Customize how queries are broken into tokens, with custom separators and dictionaries |
AI-powered search
Semantic and hybrid search using vector embeddings for meaning-based results.| Feature | Description |
|---|---|
| Hybrid search | Combine keyword and semantic search |
| Embedders | OpenAI, Hugging Face, Cohere, Mistral, Voyage, Gemini, Cloudflare, Ollama, and custom REST |
| Similar documents | Find related content automatically |
| Image search | Search images with multimodal embeddings |
| Multi-embedder | Multiple embedding models on the same document |
| User-provided vectors | Bring your own pre-generated embeddings |
Conversational search
Build chat interfaces powered by your search data with LLM integration.| Feature | Description |
|---|---|
| Chat completions | RAG-powered conversational search |
| LLM providers | OpenAI, Azure OpenAI, Mistral, Google Gemini, vLLM, and custom providers |
| Streaming responses | Stream chat responses in real time |
| Chat workspaces | Configure chat settings per index |
Personalization
| Feature | Description |
|---|---|
| User preference ranking | Re-rank results based on individual user preferences using AI embeddings |
| Real-time adaptation | Adjust results as user preferences evolve |
Filtering and faceting
Refine search results with powerful filters and build faceted navigation.| Feature | Description |
|---|---|
| Filters | Filter by any attribute with complex expressions (AND/OR) |
| Facets | Build faceted navigation interfaces |
| Facet types | AND/OR operators, numeric, boolean, and date facets |
| Sorting | Sort results by any field |
| Geo search | Geo radius, bounding box, geo sorting, and distance calculation |
Multi-search and federation
Query multiple indexes in a single request for complex search scenarios.| Feature | Description |
|---|---|
| Multi-search | Query multiple indexes at once |
| Federated search | Merge results from multiple sources with configurable weights |
Scaling
Scale Meilisearch horizontally across multiple instances or optimize resource usage on a single node.| Feature | Description |
|---|---|
| Sharding | Distribute documents across multiple instances for horizontal scaling |
| Replication | Replicate data across multiple instances for high availability |
| Remote federation | Federate search across multiple Meilisearch instances |
| Memory mapping | Efficient memory usage through memory-mapped storage |
Database
Document database
| Feature | Description |
|---|---|
| Schemaless | Index documents without a predefined schema |
| Documents | Add, replace, update, and delete documents |
| Delete by filter | Delete documents matching a filter expression |
| Update by function | Partial updates to documents using functions |
| Searchable attributes | Configure which fields are searchable and their priority |
| Displayed attributes | Control which fields are returned in results |
| Filterable attributes | Define which fields can be used in filters |
| Sortable attributes | Define which fields can be used for sorting |
| Index swap | Swap indexes to perform updates without downtime |
Vector database
| Feature | Description |
|---|---|
| Similar documents | Find semantically similar documents using vector embeddings |
| Binary quantization | Compress vectors to save storage |
| DiskANN | Disk-based approximate nearest neighbors for large datasets |
| Auto embedding | Automatically generate embeddings without manual input |
Platform features
Security
Protect your data with API keys and multi-tenant access control.| Feature | Description |
|---|---|
| API keys | Admin, search, and chat key types for different access levels |
| Tenant tokens | Secure multi-tenant applications with document-level access control |
| Search rules | Restrict which documents users can access |
Tasks and monitoring
Monitor indexing operations and receive notifications.| Feature | Description |
|---|---|
| Task management | Track and manage async operations |
| Batches | Automatic task batching for efficient processing |
| Webhooks | Get notified when tasks complete |
| Diff indexing | Only index differences between datasets |
Analytics (Cloud)
Track search behavior and optimize relevancy with built-in analytics.| Feature | Description |
|---|---|
| Search analytics | Monitor search patterns, no-result rates, and top queries |
| Click tracking | Track which results users engage with |
| Metrics reference | Click-through rate, conversion rate, average click position |
| Monitoring | Search latency, indexing latency, bandwidth, and API health |
Pagination
| Feature | Description |
|---|---|
| Pagination | Offset/limit and cursor-based pagination |
Language support
Meilisearch provides optimized support for many languages:- Latin-based languages: English, French, Spanish, German, Italian, Portuguese, etc.
- CJK: Chinese, Japanese, Korean with specialized tokenization
- RTL languages: Hebrew, Arabic
- Others: Thai, Greek, and more
Cloud features
These features are available exclusively on Meilisearch Cloud.| Feature | Description |
|---|---|
| Crawler | Crawl web pages with JS rendering, DocSearch mode, and schema extraction |
| Search preview | Visual search interface with filtering, sorting, and document CRUD |
| Teams | Organize projects and members into team workspaces |
| Enterprise SSO/SCIM | SAML 2.0 SSO and automated user provisioning |
| Autoscale disk | Automatically scale storage as data grows |
| Automatic backups | Scheduled backups for data safety |
| One-click upgrade | Upgrade Meilisearch version with a single click |
| Up to 99.999% SLA | Industry-leading service level agreement |
Self-hosting
| Feature | Description |
|---|---|
| Configuration | CLI flags and environment variables |
| Snapshots | Full binary copies for fast restoration |
| Dumps | Portable JSON exports for migration |
| Master key | Secure your instance with a master key |
Integration options
| Option | Description |
|---|---|
| REST API | Direct HTTP integration |
| Official SDKs | 10+ language SDKs |
| Frameworks | Laravel, Rails, Strapi, Symfony |