Qdrant · Open-source vector database

Qdrant

Open-source vector search engine written in Rust, available self-hosted or as a managed cloud.

RAG & Vector Databases AI Deployment & Monitoring

Provider: Qdrant
Latest confirmed model: Not applicable / not verified
Pricing: Open source
Free plan: Yes
API: Yes · free quota
Status: Generally available
Last verified: Oct 19, 2026

Visit official website Pricing page API docs

What is Qdrant?

Qdrant is a vector database that stores embeddings with attached metadata and returns the closest matches, with filtering. The core engine is open source under Apache 2.0, so you can run it yourself in Docker or Kubernetes. Qdrant Cloud offers a free single-node cluster, usage-based managed clusters, and hybrid or private deployments.

Fast similarity search with filters for RAG and search apps, with the option to keep data on your own infrastructure.

Why use Qdrant?

Practical benefits

Apache 2.0 open source
Self-host or managed cloud
Free-forever cloud cluster for testing
Hybrid Cloud for running in your own infrastructure

Who it suits

Developers building RAG
Teams with data-residency needs
Engineers who prefer open source

When another tool may fit better

Teams wanting zero operational work and fixed monthly prices
Non-technical users

Learning curve: Some experience helps
Skills needed: Programming; DevOps for self-hosting.

Example workflows

Run Qdrant in Docker locally, create a collection, upsert vectors with payloads, and search with filters

What can it do?

Open-source engineApache 2.0, written in Rust.
Managed Cloud free tierSingle node with 0.5 vCPU, 1 GB RAM and 4 GB disk, free forever.
Standard tierHourly billing for vCPU, memory, storage, backups and paid inference tokens.
Hybrid Cloud Depends on planManaged clusters on your own infrastructure.
Private Cloud Depends on planDedicated isolated deployment with custom SLAs.
Cloud inference LimitedFree cloud inference with selected models on the free tier.

Good for

RAG retrieval
Semantic and hybrid search
Recommendations
Anomaly detection

Where it runs

DockerKubernetesQdrant CloudLinuxmacOSWindows

Pricing and plans

Open source The engine is free and open source. Qdrant Cloud has a free-forever single-node cluster; Standard is billed hourly by resources used; Premium, Hybrid and Private Cloud pricing is on request.

Plan	Price	Billing	Free trial / tier	Usage limits	Main inclusions	Model access	API access	Restrictions
Open source (self-hosted)	Free	per month	Free tier	Limited by your hardware	Full engine	—	Yes	You operate it
Cloud Free Tier	Free	per month	Free tier	0.5 vCPU, 1 GB RAM, 4 GB disk, single node	Free cloud inference with selected models	—	Yes	For testing and prototypes
Standard	Billed hourly for vCPU, memory, storage, backups and inference tokens	usage-based	—	—	Production clusters	—	Yes	Subscription via Marketplace
Premium / Hybrid / Private Cloud	Contact provider	usage-based	—	Minimum spend for Premium	Custom SLAs Run in your infrastructure	—	Yes	—

Prices in USD · Global · Taxes may apply · Verified Oct 19, 2026

Official pricing page (opens in a new tab)

API and developer access

Official API: Yes
Documentation: API docs (opens in a new tab)
Authentication: API key (cloud)
Pricing model: Hourly compute and storage (cloud)
Free API quota: Free-forever cloud cluster (0.5 vCPU, 1 GB RAM, 4 GB disk)
Rate limits: Not documented
Context / input limits: Not documented
SDKs: Not documented
Streaming: Not documented
Batch processing: Not documented
Structured output: Not documented
Tool / function calling: Not documented
Commercial use: Apache 2.0 for the open-source engine
Data retention & privacy: See the provider privacy terms

A free Qdrant account is not the same thing as free API access. The API allowance above is the provider's documented developer quota.

Verified Oct 19, 2026

Advantages and disadvantages

Advantages

Permissive open-source licence
Flexible deployment options
Free cloud tier

Disadvantages

Managed pricing is usage-based without published dollar amounts on the pricing page
Self-hosting requires operational effort
Free cluster is small

Privacy: Self-hosting and Hybrid/Private Cloud keep data in your own environment.

Real-world examples

A developer testing locally

Run the Qdrant Docker image
Create a collection and upload embeddings
Query with metadata filters

free plan works

A startup prototyping in the cloud

Create a free-tier cluster in Qdrant Cloud
Connect with an API key
Upgrade to Standard when traffic grows

free plan works

A regulated company keeping data in-house

Choose Hybrid or Private Cloud
Deploy in your own environment
Let Qdrant manage cluster operations

needs paid plan

Qdrant alternatives

Pinecone

Fully managed vector database for semantic search and retrieval-augmented generation.

has a free plan · offers an API

Qdrant vs Pinecone

Chroma

Open-source embedding database that runs in your app or as Chroma Cloud.

has a free plan · offers an API · open source / open-weight

Qdrant vs Chroma

LlamaIndex

Open-source framework for connecting LLMs to your data, plus the commercial LlamaParse document platform.

has a free plan · offers an API · open source / open-weight

Qdrant vs LlamaIndex

Compare Qdrant with these side by side

Frequently asked questions

Is Qdrant free?

Yes. The engine is open source under Apache 2.0 and free to self-host, and Qdrant Cloud has a free-forever single-node cluster.

What does the free cloud cluster include?

One node with 0.5 vCPU, 1 GB RAM and 4 GB disk, intended for testing and prototypes.

How is Qdrant Cloud billed?

The Standard tier is billed hourly for the vCPU, memory and storage your clusters use, plus backups and paid inference tokens.

Can I use Qdrant commercially for free?

The Apache 2.0 licence allows commercial use of the self-hosted engine.

Can Qdrant run in my own cloud?

Yes, via self-hosting or Hybrid Cloud, where Qdrant manages clusters on your infrastructure.