# Vector Search with Manticore Search

Achieve Next-Level Search Accuracy with Vector Search in Manticore.

## What is Vector Search?

Vector Search enables finding items by comparing their vector representations, making it ideal for tasks such as semantic search, recommendation systems, and image similarity detection. Manticore Search offers built-in support for vector search, making it simple to integrate this powerful feature into your applications.

## When to use Vector Search?

- Implementing semantic search for natural language queries
- Building recommendation systems for products or content
- Searching for similar images or audio files
- Detecting duplicate or near-duplicate content
- Analyzing similarities in user behavior or preferences
- Implementing facial recognition systems
- Clustering similar documents or data points
- Enhancing search results with contextual understanding
- Implementing anomaly detection in time-series data
- Performing multi-modal search combining text and images


## Why Manticore Search is good for Vector Search

- Manticore Search offers native support for vector search, eliminating the need for additional specialized tools.
- It supports multiple distance metrics like cosine, dot product, and L2 distance for flexible similarity calculations.
- Manticore's vector search works well alongside native hybrid search when you need both semantic similarity and exact keyword matching.
- The implementation is highly optimized for performance, even with large-scale vector datasets.


## Pros

- Native support for vector search without additional tools
- Multiple distance metrics available for flexibility
- Works well alongside native hybrid search
- Highly optimized for performance with large vector datasets
- Seamless integration with existing Manticore Search features
- Supports both SQL and JSON interfaces


## Cons

- Requires pre-computed vector representations of your data
- May require more storage space compared to traditional text-only indexes
- Performance can degrade with very high-dimensional vectors


## How to get started

### Install Manticore Search

- Visit the official Manticore Search website: https://manticoresearch.com/
- Follow the [installation instructions](/install/) for your operating system
- Alternatively, use Docker: `docker pull manticoresearch/manticore`


### Set up your Manticore Search index with vector fields

- Create your table schema including vector fields
- Specify the dimension of your vectors and the distance metric to use
- Index your data, including both text and vector representations


### Implement vector search functionality

- Use the `KNN` clause with vector fields to perform similarity searches
- Use Manticore's hybrid search when you need to combine vector search with full-text search
- Implement a client-side HTTP request to query Manticore Search


### Fine-tune your vector search

- Experiment with different distance metrics to find the best fit for your data
- Use query-time parameters to adjust search behavior
- Implement post-processing steps if needed, such as re-ranking results


### Enjoy Vector Search

- Experience powerful similarity-based search with Manticore Search
- Feel free to [create an issue](https://github.com/manticoresoftware/manticoresearch/) if you encounter any problems
- Consider our [professional services](/services/) for advanced implementations


## Resources

- [Vector Search Guide](https://manual.manticoresearch.com/Searching/KNN)
- [Vector Search demo](https://github.manticoresearch.com/manticoresoftware/manticoresearch?query=improve+performance&filters%5Bcomment%5D%5B%5D=28798446&filters%5Bcommon%5D%5Brepo_id%5D%5B%5D=95614931&sort=&search=keyword-search-fuzzy-layouts)
- [Blog post about Vector Search in Manticore](/blog/vector-search/)
- [Blog post about Vector Search in old and modern databases](/blog/vector-search-in-databases/)
- [Youtube Video about Vector Search in old and modern databases](https://www.youtube.com/watch?v=sHwUgxztcqs&ab_channel=Manticore)
- [Hybrid Search documentation](https://manual.manticoresearch.com/Searching/Hybrid_search)


## Learn more about other use cases

Do not stop here when learning when you need **AI Database** and how **Manticore Search** can help you. There are many other use cases that you can explore.


## Install Manticore Search to use Vector Search

Try Manticore Search for implementing advanced Vector Search in your applications today!

[Install Now](/install/)