Qdrant

Qdrant is a high-performance, open-source vector database and search engine designed for storing, managing, and querying high-dimensional vector embeddings. Headquartered in Berlin, Qdrant's core technology is written in Rust, which enables fast and reliable performance even under heavy loads. The database supports various distance metrics including cosine similarity, dot product, and Euclidean distance for vector searches. Qdrant utilizes the Hierarchical Navigable Small World (HNSW) algorithm for efficient indexing and retrieval of vector data.

Qdrant's vector database allows developers to store and search unstructured data such as images, audio, video, and text in the form of vector embeddings. It provides features like advanced filtering support, the ability to handle billions of vectors, and efficient indexing.

In April 2023, Qdrant introduced a new compression technology called “binary quantization” (BQ), which can reduce memory consumption by up to 32x and enhance retrieval speeds by around 40x. The company offers an open-source version under the Apache 2.0 license and a managed cloud service. Qdrant's cloud platform is available on Microsoft Azure, along with existing support for AWS and Google Cloud Platform.

In April 2024, Quadrant introduced “Qdrant Hybrid Cloud” to offer users a vector database that they can deploy in various environments, covering cloud, on-premise, or Edge locations.

Key customers and partnerships

Qdrant’s clientele includes notable firms including Deloitte, Accenture, and X (formerly Twitter).

In January 2024, Qdrant announced a partnership with Microsoft to make its vector database available on the Azure cloud platform, enabling seamless integration for enterprise AI applications.