Qdrant Framework Integrations & SDKs 2026
Qdrant ships dedicated pages for 41 frameworks and six official SDKs, across Python, JavaScript, Rust, Go, .NET and Java. No community-only gap for major languages.
Qdrant Framework Integrations verdict
Qdrant ships dedicated integration pages for 41 frameworks, including LangChain, LlamaIndex, Haystack, CrewAI, DSPy and AutoGen.
Six first-party SDKs cover Python, JavaScript, Rust, Go, .NET and Java, at v1.18.x. There is no community-only gap for major languages.
Pick Qdrant if your stack is Python, Rust or Go and you want native integration with every major RAG framework plus multi-agent orchestrators. Java and .NET teams get official clients. The hybrid retrieval mode, dense plus sparse, is first-class but requires configuring a sparse embedder.
- The LangChain integration requires the langchain-qdrant package, not langchain-community, and the legacy import path is deprecated.
- Hybrid dense-plus-sparse retrieval requires providing a sparse embedding implementation, such as FastEmbedSparse. It is not automatic.
- The Java and .NET clients are official but lag Python by minor patch versions, at 1.18.3 and 1.18.1 against Python's 1.18.0.
- Framework integrations
- 41
- Official SDK languages
- 6
- Python client version
- v1.18.0
- Hybrid retrieval modes
- Dense + Sparse + Hybrid
- Agent frameworks
- CrewAI, AutoGen, DSPy, LangGraph
Framework and agent-orchestration integrations
| Integration | Type | Capabilities | Setup |
|---|---|---|---|
| LangChain | Native RAG | QdrantVectorStore · dense/sparse/hybrid retrieval · metadata filtering | Low |
| LlamaIndex | Native RAG | QdrantVectorStore · VectorStoreIndex · document ingestion at scale | Low |
| Haystack | Native RAG | QdrantDocumentStore · scalar quantization config · production pipelines | Low |
| CrewAI | Agent memory | short-term memory · entity memory · multi-agent collaboration | Medium |
| AutoGen | Agent RAG | multi-agent RAG · document retrieval · LLM reasoning | Medium |
| DSPy | Retrieval RM | retrieval module (RM) · prompt optimization · fine-tuning pipeline | Medium |
| LangGraph | Stateful agents | stateful multi-actor apps · Python and JS · graph-based workflows | Medium |
| Agno | Agent framework | fast multi-agent runtime · control plane · vector memory | Medium |
Find your integration path into Qdrant
- LangChain is a native rag integration (QdrantVectorStore, dense/sparse/hybrid retrieval, metadata filtering), low setup effort.
- Official Python SDK v1.18.0 targets API .
- Install:
pip install langchain-qdrant
Based on Qdrant's documented integrations and official SDK matrix.
LangChain + Qdrant quickstart
from langchain_qdrant import QdrantVectorStore
from langchain_openai import OpenAIEmbeddings
embeddings = OpenAIEmbeddings()
# Create a new collection from documents
doc_store = QdrantVectorStore.from_documents(
docs, embeddings, url="https://<cluster>.qdrant.io",
api_key="<api-key>", collection_name="knowledge_base"
)
# Or connect to an existing collection
store = QdrantVectorStore.from_existing_collection(
embedding=embeddings, url="https://<cluster>.qdrant.io",
api_key="<api-key>", collection_name="knowledge_base"
)
# Retrieve with optional metadata filter
hits = store.similarity_search("how does billing work?", k=4)Official SDKs and client libraries
| Language | Package | Version | Install |
|---|---|---|---|
| Python | qdrant-client | v1.18.0 | pip install qdrant-client[fastembed] |
| JavaScript | @qdrant/js-client-rest | v1.18.0 | npm install @qdrant/js-client-rest |
| Rust | qdrant-client (crates.io) | v1.18.0 | cargo add qdrant-client |
| Go | go-client | v1.18.2 | go get github.com/qdrant/go-client |
| .NET | Qdrant.Client (NuGet) | v1.18.1 | dotnet add package Qdrant.Client |
| Java | qdrant-java-client | v1.18.3 | see Maven Central |
Qdrant LangChain integration detail
- Install via langchain-qdrant package, not the legacy langchain-community package
- Primary class is QdrantVectorStore, supporting from_texts, from_documents, and from_existing_collection factories
- Supports RetrievalMode.DENSE (default), RetrievalMode.SPARSE, and RetrievalMode.HYBRID via retrieval_mode parameter
- FastEmbedSparse is available for sparse vector operations without external embedding service
- similarity_search accepts k and metadata filter parameters, returning ranked documents
- Can connect to local in-memory instance (location=':memory:') or cloud cluster (url + api_key)
Qdrant Framework Integrations FAQ
Which RAG frameworks does Qdrant support natively?
LangChain, LlamaIndex and Haystack all have dedicated integration pages in Qdrant's docs. The langchain-qdrant package exposes QdrantVectorStore with dense, sparse and hybrid retrieval modes. LlamaIndex wraps it as QdrantVectorStore fed into VectorStoreIndex. Haystack uses QdrantDocumentStore with configurable quantization.
Which SDK languages does Qdrant support officially?
Six: Python at v1.18.0, JavaScript or TypeScript at v1.18.0, Rust at v1.18.0, Go at v1.18.2, .NET at v1.18.1 and Java at v1.18.3. All are maintained by Qdrant and use a gRPC or REST transport.
How do I install Qdrant with LangChain?
Install langchain-qdrant, not langchain-community. The main class is QdrantVectorStore. Use from_documents to build a new collection or from_existing_collection to connect to one already in Qdrant. Set retrieval_mode to DENSE, SPARSE or HYBRID as needed.
Does Qdrant support agent frameworks beyond LangChain?
Yes. Qdrant has dedicated pages for CrewAI, covering agent short-term and entity memory, AutoGen for multi-agent RAG, DSPy as a retrieval module, and LangGraph for stateful multi-actor apps. In total, 41 framework integrations are documented.
How many total framework integrations does Qdrant have?
41 in total. They span RAG orchestration like LangChain, LlamaIndex and Haystack, and multi-agent systems like CrewAI, AutoGen, LangGraph and SmolAgents. They also cover evaluation tools like DeepEval and HoneyHive, and data processing pipelines like Cognee, Sycamore and Vanna AI.
Sources & verification
| Source | What was checked | Last checked |
|---|---|---|
| Qdrant Official | Official product page | July 10, 2026 |
| Docs Developer docs | Qdrant Client Latest | July 10, 2026 |
| GitHub Go Client Releases | Go Client Releases | July 10, 2026 |
| GitHub Qdrant Client Releases | Qdrant Client Releases | July 10, 2026 |
| GitHub Qdrant Js Releases | Qdrant Js Releases | July 10, 2026 |
| Nuget Packages Qdrant.Client | Packages Qdrant.client | July 10, 2026 |
| Qdrant Documentation Frameworks | Documentation Frameworks | July 10, 2026 |
Every fact on this Qdrant page is tied to a named source and a verification date. Freshness-sensitive figures trace to the sources above; verify against the vendor before relying on them.
Explore Qdrant
Every page on Qdrant in one place, you are on framework integrations.
Snapshot, score and verdict
You are here
Latency, throughput, uptime and behaviour under scale
Every tier and the entry price
Compared and ranked vs peers
Price and feature change history
Browse the full Vector Databases category