# AI Search

Create fully managed RAG pipelines for your AI applications

> Links below point directly to Markdown versions of each page. Any page can also be retrieved as Markdown by sending an `Accept: text/markdown` header to the page's URL without the `index.md` suffix (for example, `curl -H "Accept: text/markdown" https://docs.ahq.lat/ai-search/`).
>
> For other Cloudflare products, see the [Cloudflare documentation directory](https://docs.ahq.lat/llms.txt).

## Overview

- [Overview](https://docs.ahq.lat/ai-search/index.md): Cloudflare AI Search is a managed search service. Index your content and query it with natural language from a Workers binding, REST API, or MCP server.

## Get started

- [Get started](https://docs.ahq.lat/ai-search/get-started/index.md): Create fully-managed, retrieval-augmented generation pipelines with Cloudflare AI Search.
- [REST API](https://docs.ahq.lat/ai-search/get-started/api/index.md): Create AI Search instances programmatically using the REST API.
- [Dashboard](https://docs.ahq.lat/ai-search/get-started/dashboard/index.md): Create and configure AI Search using the Cloudflare dashboard.
- [Wrangler commands](https://docs.ahq.lat/ai-search/get-started/wrangler/index.md): Create and manage AI Search instances from the command line.

## Wrangler commands

- [Wrangler commands](https://docs.ahq.lat/ai-search/wrangler-commands/index.md): Manage AI Search instances from the command line using Wrangler.

## Configuration

- [Configuration](https://docs.ahq.lat/ai-search/configuration/index.md): Customize how your AI Search instance indexes data, retrieves results, and generates responses.
- [Data source](https://docs.ahq.lat/ai-search/configuration/data-source/index.md): Connect a website, R2 bucket, or upload files directly to your AI Search instance for indexing.
- [Built-in storage](https://docs.ahq.lat/ai-search/configuration/data-source/built-in-storage/index.md): Upload files directly to an AI Search instance using built-in storage powered by R2 and Vectorize.
- [R2](https://docs.ahq.lat/ai-search/configuration/data-source/r2/index.md): Connect a Cloudflare R2 bucket as a data source to index stored documents with AI Search.
- [Website](https://docs.ahq.lat/ai-search/configuration/data-source/website/index.md): Connect a domain you own as a data source so AI Search can crawl and index your website pages.
- [Chunking](https://docs.ahq.lat/ai-search/configuration/indexing/chunking/index.md): Configure how AI Search splits content into chunks for embedding and retrieval.
- [Hybrid search](https://docs.ahq.lat/ai-search/configuration/indexing/hybrid-search/index.md): Combine vector and keyword search in AI Search for broader, more accurate retrieval results.
- [Keyword search](https://docs.ahq.lat/ai-search/configuration/indexing/keyword-search/index.md): Enable BM25 keyword search in AI Search to match documents containing exact query terms.
- [Metadata](https://docs.ahq.lat/ai-search/configuration/indexing/metadata/index.md): Use metadata attributes and custom schemas in AI Search to filter and contextualize search results.
- [Path filtering](https://docs.ahq.lat/ai-search/configuration/indexing/path-filtering/index.md): Control which files or URLs AI Search indexes by defining include and exclude path patterns.
- [Service API token](https://docs.ahq.lat/ai-search/configuration/indexing/service-api-token/index.md): Create a service API token to grant AI Search read access to R2 buckets for indexing.
- [Syncing](https://docs.ahq.lat/ai-search/configuration/indexing/syncing/index.md): Understand how AI Search automatically syncs and indexes content from connected data sources.
- [Vector search](https://docs.ahq.lat/ai-search/configuration/indexing/vector-search/index.md): Configure vector search in AI Search to find semantically similar content using embeddings.
- [Models](https://docs.ahq.lat/ai-search/configuration/models/index.md): Configure which AI models AI Search uses for embedding, generation, reranking, and query rewriting.
- [Supported models](https://docs.ahq.lat/ai-search/configuration/models/supported-models/index.md): View all AI models supported by AI Search, including text generation, embedding, and reranking models.
- [Relevance boosting](https://docs.ahq.lat/ai-search/configuration/retrieval/boosting/index.md): Bias AI Search results toward documents with specific metadata using relevance boosting.
- [Similarity cache](https://docs.ahq.lat/ai-search/configuration/retrieval/cache/index.md): Speed up AI Search responses by caching and reusing answers for semantically similar queries.
- [UI snippets](https://docs.ahq.lat/ai-search/configuration/retrieval/embed-search-snippets/index.md): Add AI Search to your website using pre-built, customizable web components for search and chat.
- [Public endpoint settings](https://docs.ahq.lat/ai-search/configuration/retrieval/public-endpoint/index.md): Expose AI Search instances through public MCP, chat, and search endpoints without authentication.
- [Query rewriting](https://docs.ahq.lat/ai-search/configuration/retrieval/query-rewriting/index.md): Improve AI Search retrieval quality by enabling query rewriting to rephrase user queries.
- [Reranking](https://docs.ahq.lat/ai-search/configuration/retrieval/reranking/index.md): Enable reranking in AI Search to reorder retrieved results by semantic relevance.
- [Result filtering](https://docs.ahq.lat/ai-search/configuration/retrieval/result-filtering/index.md): Control AI Search result count and minimum score thresholds for returned results.
- [System prompt](https://docs.ahq.lat/ai-search/configuration/retrieval/system-prompt/index.md): Guide AI Search query rewriting and response generation behavior with custom system prompts.

## REST API

- [REST API](https://docs.ahq.lat/api/resources/ai_search/index.md): Access the AI Search REST API reference for managing instances, items, and search operations.

## api

- [REST API](https://docs.ahq.lat/ai-search/api/instances/rest-api/index.md): Manage AI Search instances and sync jobs over HTTP using the Instances REST API.
- [Workers binding](https://docs.ahq.lat/ai-search/api/instances/workers-binding/index.md): Manage AI Search instances from a Cloudflare Worker using the Instances Workers binding.
- [REST API](https://docs.ahq.lat/ai-search/api/items/rest-api/index.md): Upload, list, and manage documents in AI Search instances using the Items REST API.
- [Workers binding](https://docs.ahq.lat/ai-search/api/items/workers-binding/index.md): Upload, list, and manage documents in AI Search instances using the Items Workers binding.
- [Metadata filter (legacy)](https://docs.ahq.lat/ai-search/api/migration/autorag-filter-format/index.md): Reference for the legacy AutoRAG metadata filter format used with the previous REST API.
- [REST API migration](https://docs.ahq.lat/ai-search/api/migration/rest-api/index.md): Migrate from the legacy AutoRAG REST API endpoints to the new AI Search API endpoints.
- [Workers binding migration](https://docs.ahq.lat/ai-search/api/migration/workers-binding/index.md): Upgrade from the legacy env.AI.autorag() binding to the new AI Search Workers bindings.
- [Workers binding (legacy)](https://docs.ahq.lat/ai-search/api/migration/workers-binding-legacy/index.md): Reference for the legacy env.AI.autorag() Workers binding used by earlier AI Search instances.
- [MCP](https://docs.ahq.lat/ai-search/api/search/mcp/index.md): Expose AI Search content to AI agents through the Model Context Protocol (MCP) endpoint.
- [Public endpoint](https://docs.ahq.lat/ai-search/api/search/public-endpoint/index.md): Integrate AI Search into public-facing applications using unauthenticated public endpoints.
- [REST API](https://docs.ahq.lat/ai-search/api/search/rest-api/index.md): Query AI Search instances over HTTP using the REST API for search and chat completions.
- [Workers binding](https://docs.ahq.lat/ai-search/api/search/workers-binding/index.md): Search and chat with AI Search instances from a Cloudflare Worker using the Workers binding.

## concepts

- [How AI Search works](https://docs.ahq.lat/ai-search/concepts/how-ai-search-works/index.md): Understand how AI Search indexes your content and retrieves results using vector and keyword search.
- [Namespaces](https://docs.ahq.lat/ai-search/concepts/namespaces/index.md): Group AI Search instances into namespaces and manage them dynamically from a Workers binding.
- [Search modes](https://docs.ahq.lat/ai-search/concepts/search-modes/index.md): Compare AI Search vector, keyword, and hybrid search modes to choose the right retrieval strategy.

## how-to

- [Bring your own generation model](https://docs.ahq.lat/ai-search/how-to/bring-your-own-generation-model/index.md): Use AI Search for retrieval while generating responses with an external model like OpenAI.
- [NLWeb](https://docs.ahq.lat/ai-search/how-to/nlweb/index.md): Deploy NLWeb with AI Search to enable conversational natural language queries on your website.
- [Build per-tenant search](https://docs.ahq.lat/ai-search/how-to/per-tenant-search/index.md): Isolate search results per tenant in AI Search using separate instances or metadata filtering.
- [Create a simple search engine](https://docs.ahq.lat/ai-search/how-to/simple-search-engine/index.md): Build a simple search engine using the AI Search Workers binding and the search method.

## platform

- [Limits & pricing](https://docs.ahq.lat/ai-search/platform/limits-pricing/index.md): View AI Search usage limits and pricing details for Free and Paid Workers plans.
- [Release note](https://docs.ahq.lat/ai-search/platform/release-note/index.md): Review recent changes to Cloudflare AI Search.