Weaviate: гибридный поиск и RAG в одной БД

# Weaviate <img alt='Weaviate logo' src='https://weaviate.io/img/site/weaviate-logo-light.png' width='148' align='right' />

[](https://github.com/weaviate/weaviate) [](https://pkg.go.dev/github.com/weaviate/weaviate) [](https://github.com/weaviate/weaviate/actions/workflows/.github/workflows/pull_requests.yaml) [](https://goreportcard.com/report/github.com/weaviate/weaviate) [](https://codecov.io/gh/weaviate/weaviate)

**Weaviate** is an open-source, cloud-native vector database that stores both objects and vectors, enabling semantic search at scale. It combines vector similarity search with keyword filtering, retrieval-augmented generation (RAG), and reranking in a single query interface. Common use cases include RAG systems, semantic and image search, recommendation engines, chatbots, and content classification.

Weaviate supports two approaches to store vectors: automatic vectorization at import using [integrated models](https://docs.weaviate.io/weaviate/model-providers) (OpenAI, Cohere, HuggingFace, and others) or direct import of [pre-computed vector embeddings](https://docs.weaviate.io/weaviate/starter-guides/custom-vectors). Production deployments benefit from built-in multi-tenancy, replication, RBAC authorization, and [many other features](#weaviate-features).

To get started quickly, have a look at one of these tutorials:

- [Quickstart - Weaviate Cloud](https://docs.weaviate.io/weaviate/quickstart) - [Quickstart - local Docker instance](https://docs.weaviate.io/weaviate/quickstart/local)

## Installation

Weaviate offers multiple installation and deployment options:

- [Docker](https://docs.weaviate.io/deploy/installation-guides/docker-installation) - [Kubernetes](https://docs.weaviate.io/deploy/installation-guides/k8s-installation) - [Weaviate Cloud](https://console.weaviate.cloud)

See the [installation docs](https://docs.weaviate.io/deploy) for more deployment options, such as [AWS](https://docs.weaviate.io/deploy/installation-guides/aws-marketplace) and [GCP](https://docs.weaviate.io/deploy/installation-guides/gcp-marketplace).

## Getting started

You can easily start Weaviate and a local vector embedding model with [Docker](https://docs.docker.com/desktop/). Create a `docker-compose.yml` file:

```yml services: weaviate: image: cr.weaviate.io/semitechnologies/weaviate:1.36.0 ports: - "8080:8080" - "50051:50051" environment: ENABLE_MODULES: text2vec-model2vec MODEL2VEC_INFERENCE_API: http://text2vec-model2vec:8080

# A lightweight embedding model that will generate vectors from objects during import text2vec-model2vec: image: cr.weaviate.io/semitechnologies/model2vec-inference:minishlab-potion-base-32M ```

Start Weaviate and the embedding service with:

```bash docker compose up -d ```

Install the Python client (or use another [client library](#client-libraries-and-apis)):

```bash pip install -U weaviate-client ```

The following Python example shows how easy it is to populate a Weaviate database with data, create vector embeddings and perform semantic search:

```python import weaviate from weaviate.classes.config import Configure, DataType, Property

# Connect to Weaviate client = weaviate.connect_to_local()

# Create a collection client.collections.create( name="Article", properties=[Property(name="content", data_type=DataType.TEXT)], vector_config=Configure.Vectors.text2vec_model2vec(), # Use a vectorizer to generate embeddings during import # vector_config=Configure.Vectors.self_provided() # If you want to import your own pre-generated embeddings )

# Insert objects and generate embeddings articles = client.collections.get("Article") articles.data.insert_many( [ {"content": "Vector databases enable semantic search"}, {"content": "Machine learning models generate embeddings"}, {"content": "Weaviate supports hybrid search capabilities"}, ] )

# Perform semantic search results = articles.query.near_text(query="Search objects by meaning", limit=1) print(results.objects[0])

client.close() ```

This example uses the `Model2Vec` vectorizer, but you can choose any other [embedding model provider](https://docs.weaviate.io/weaviate/model-providers) or [bring your own pre-generated vectors](https://docs.weaviate.io/weaviate/starter-guides/custom-vectors).

## Client libraries and APIs

Weaviate provides client libraries for several programming languages:

- [Python](https://docs.weaviate.io/weaviate/client-libraries/python) - [JavaScript/TypeScript](https://docs.weaviate.io/weaviate/client-libraries/typescript) - [Java](https://docs.weaviate.io/weaviate/client-libraries/java) - [Go](https://docs.weaviate.io/weaviate/client-libraries/go) - [C#/.NET](https://docs.weaviate.io/weaviate/client-libraries/csharp)

There are also additional [community-maintained libraries](https://docs.weaviate.io/weaviate/client-libraries/community).

Weaviate exposes [REST API](https://docs.weaviate.io/weaviate/api/rest), [gRPC API](https://docs.weaviate.io/weaviate/api/grpc), and [GraphQL API](https://docs.weaviate.io/weaviate/api/graphql) to communicate with the database server.

## Weaviate features

These features enable you to build AI-powered applications:

- **⚡ Fast Search Performance**: Perform complex semantic [searches](https://docs.weaviate.io/weaviate/search/similarity) over billions of vectors in milliseconds. Weaviate's architecture is built in Go for speed and reliability, ensuring your AI applications are highly responsive even under heavy load. See our [ANN benchmarks](https://docs.weaviate.io/weaviate/benchmarks/ann) for more info.

- **🔌 Flexible Vectorization**: Seamlessly vectorize data at import time with [integrated vectorizers](https://docs.weaviate.io/weaviate/model-providers) from OpenAI, Cohere, HuggingFace, Google, and more. Or you can import [your own vector embeddings](https://docs.weaviate.io/weaviate/starter-guides/custom-vectors).

- **🔍 Advanced Hybrid & Image Search**: Combine the power of semantic search with traditional [keyword (BM25) search](https://docs.weaviate.io/weaviate/search/bm25), [image search](https://docs.weaviate.io/weaviate/search/image) and [advanced filtering](https://docs.weaviate.io/weaviate/search/filters) to get the best results with a single API call.

- **🤖 Integrated RAG & Reranking**: Go beyond simple retrieval with built-in [generative search (RAG)](https://docs.weaviate.io/weaviate/search/generative) and [reranking](https://docs.weaviate.io/weaviate/search/rerank) capabilities. Power sophisticated Q&A systems, chatbots, and summarizers directly from your database without additional tooling.

- **📈 Production-Ready & Scalable**: Weaviate is built for mission-critical applications. Go from rapid prototyping to production at scale with native support for [horizontal scaling](https://docs.weaviate.io/deploy/configuration/horizontal-scaling), [multi-tenancy](https://docs.weaviate.io/weaviate/manage-collections/multi-tenancy), [replication](https://docs.weaviate.io/deploy/configuration/replication), and fine-grained [role-based access control (RBAC)](https://docs.weaviate.io/weaviate/configuration/rbac).

- **💰 Cost-Efficient Operations**: Radically lower resource consumption and operational costs with built-in [vector compression](https://docs.weaviate.io/weaviate/configuration/compression). Vector quantization and multi-vector encoding reduce memory usage with minimal impact on search performance.

- **⏱️ Object TTL**: Automatically expire and remove stale data with configurable [time-to-live](https://docs.weaviate.io/weaviate/manage-collections/time-to-live) settings per collection, with full RBAC and multi-tenancy support.

For a complete list of all functionalities, visit the [official Weaviate documentation](https://docs.weaviate.io).

## Useful resources

### AI Agent Skills

[Weaviate Agent Skills](https://github.com/weaviate/agent-skills) is a collection of skills for AI coding agents (Claude Code, Cursor, GitHub Copilot, and others) that enable them to work with Weaviate more accurately and efficiently. Skills cover searching, querying, collection management, data import, and full application blueprints (RAG, agentic RAG, chatbots, and more).

Install with:

```bash npx skills add weaviate/agent-skills ```

### Demo projects & recipes

These demos are working applications that highlight some of Weaviate's capabilities. Their source code is available on GitHub.

- [Elysia](https://elysia.weaviate.io) ([GitHub](https://github.com/weaviate/elysia)): Elysia is a decision tree based agentic system which intelligently decides what tools to use, what results have been obtained, whether it should continue the process or whether its goal has been completed. - [Verba](https://weaviate.io/blog/verba-open-source-rag-app) ([GitHub](https://github.com/weaviate/verba)): A community-driven open-source application designed to offer an end-to-end, streamlined, and user-friendly interface for Retrieval-Augmented Generation (RAG) out of the box. - [Healthsearch](https://weaviate.io/blog/healthsearch-demo) ([GitHub](https://github.com/weaviate/healthsearch-demo)): An open-source project aimed at showcasing the potential of leveraging user-written reviews and queries to retrieve supplement products based on specific health effects. - Awesome-Moviate ([GitHub](https://github.com/weaviate-tutorials/awesome-moviate)): A movie search and recommendation engine that allows keyword-based (BM25), semantic, and hybrid searches.

We also maintain extensive repositories of **Jupyter Notebooks** and **TypeScript code snippets** that cover how to use Weaviate features and integrations:

- [Weaviate Python Recipes](https://github.com/weaviate/recipes/) - [Weaviate TypeScript Recipes](https://github.com/weaviate/recipes-ts/)

### Blog posts

- [What is a Vector Database](https://weaviate.io/blog/what-is-a-vector-database) - [What is Vector Search](https://weaviate.io/blog/vector-search-explained) - [What is Hybrid Search](https://weaviate.io/blog/hybrid-search-explained) - [How to Choose an Embedding Model](https://weaviate.io/blog/how-to-choose-an-embedding-model) - [What is RAG](https://weaviate.io/blog/introduction-to-rag) - [RAG Evaluation](https://weaviate.io/blog/rag-evaluation) - [Advanced RAG Techniques](https://weaviate.io/blog/advanced-rag) - [What is Multimodal RAG](https://weaviate.io/blog/multimodal-rag) - [What is Agentic RAG](https://weaviate.io/blog/what-is-agentic-rag) - [What is Graph RAG](https://weaviate.io/blog/graph-rag) - [Overview of Late Interaction Models](https://weaviate.io/blog/late-interaction-overview)

### Integrations

Weaviate integrates with many external services:

| Category | Description | Integrations | | ------------------------------------------------------------------------------------------ | ---------------------------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | **[Cloud Hyperscalers](https://docs.weaviate.io/integrations/cloud-hyperscalers)** | Large-scale computing and storage | [AWS](https://docs.weaviate.io/integrations/cloud-hyperscalers/aws), [Google](https://docs.weaviate.io/integrations/cloud-hyperscalers/google) | | **[Compute Infrastructure](https://docs.weaviate.io/integrations/compute-infrastructure)** | Run and scale containerized applications | [Modal](https://docs.weaviate.io/integrations/compute-infrastructure/modal), [Replicate](https://docs.weaviate.io/integrations/compute-infrastructure/replicate), [Replicated](https://docs.weaviate.io/integrations/compute-infrastructure/replicated) | | **[Data Platforms](https://docs.weaviate.io/integrations/data-platforms)** | Data ingestion and web scraping | [Airbyte](https://docs.weaviate.io/integrations/data-platforms/airbyte), [Aryn](https://docs.weaviate.io/integrations/data-platforms/aryn), [Boomi](https://docs.weaviate.io/integrations/data-platforms/boomi), [Box](https://docs.weaviate.io/integrations/data-platforms/box), [Confluent](https://docs.weaviate.io/integrations/data-platforms/confluent), [Astronomer](https://docs.weaviate.io/integrations/data-platforms/astronomer), [Context Data](https://docs.weaviate.io/integrations/data-platforms/context-data), [Databricks](https://docs.weaviate.io/integrations/data-platforms/databricks), [Firecrawl](https://docs.weaviate.io/integrations/data-platforms/firecrawl), [IBM](https://docs.weaviate.io/integrations/data-platforms/ibm), [Unstructured](https://docs.weaviate.io/integrations/data-platforms/unstructured) | | **[LLM and Agent Frameworks](https://docs.weaviate.io/integrations/llm-agent-frameworks)** | Build agents and generative AI applications | [Agno](https://docs.weaviate.io/integrations/llm-agent-frameworks/agno), [Composio](https://docs.weaviate.io/integrations/llm-agent-frameworks/composio), [CrewAI](https://docs.weaviate.io/integrations/llm-agent-frameworks/crewai), [DSPy](https://docs.weaviate.io/integrations/llm-agent-frameworks/dspy), [Dynamiq](https://docs.weaviate.io/integrations/llm-agent-frameworks/dynamiq), [Haystack](https://docs.weaviate.io/integrations/llm-agent-frameworks/haystack), [LangChain](https://docs.weaviate.io/integrations/llm-agent-frameworks/langchain), [LlamaIndex](https://docs.weaviate.io/integrations/llm-agent-frameworks/llamaindex), [N8n](https://docs.weaviate.io/integrations/llm-agent-frameworks/n8n), [Semantic Kernel](https://docs.weaviate.io/integrations/llm-agent-frameworks/semantic-kernel) | | **[Operations](https://docs.weaviate.io/integrations/operations)** | Tools for monitoring and analyzing generative AI workflows | [AIMon](https://docs.weaviate.io/integrations/operations/aimon), [Arize](https://docs.weaviate.io/integrations/operations/arize), [Cleanlab](https://docs.weaviate.io/integrations/operations/cleanlab), [Comet](https://docs.weaviate.io/integrations/operations/comet), [DeepEval](https://docs.weaviate.io/integrations/operations/deepeval), [Langtrace](https://docs.weaviate.io/integrations/operations/langtrace), [LangWatch](https://docs.weaviate.io/integrations/operations/langwatch), [Nomic](https://docs.weaviate.io/integrations/operations/nomic), [Patronus AI](https://docs.weaviate.io/integrations/operations/patronus), [Ragas](https://docs.weaviate.io/integrations/operations/ragas), [TruLens](https://docs.weaviate.io/integrations/operations/trulens), [Weights & Biases](https://docs.weaviate.io/integrations/operations/wandb) |

## Contributing

We welcome and appreciate contributions! Please see our [Contributor guide](https://docs.weaviate.io/contributor-guide) for the development setup, code style guidelines, testing requirements and the pull request process.

Join our [Community forum](https://forum.weaviate.io/) to discuss ideas and get help.

## License

BSD 3-Clause License. See [LICENSE](./LICENSE) for details.