Qdrant is the one project in our catalog listed as a replacement for Pinecone. It is databases software with 35.3K GitHub stars. Below is how it measures against the questions people usually ask first.
A fast vector database for semantic search and AI retrieval.
Qdrant turns embeddings from AI models into searchable data. It is the storage and retrieval layer behind many semantic search, recommendation and retrieval-augmented generation systems, built for speed and reliability under load.
Store and search embeddings with attached payloads through a convenient API.
Combine vector search with rich filtering for semantic and faceted search.
Run it from a container, in Qdrant Cloud, or embedded on edge devices with Qdrant Edge.
What people use it for
Give an AI assistant your documents
Embed a client's documents, store them in Qdrant, and retrieve the most relevant passages for each question in a retrieval-augmented generation pipeline.
Build semantic product search
Search a catalogue by meaning rather than exact keywords, and narrow results with filters on price, category or availability.
Power recommendations
Find items similar to those a user liked by comparing their vectors, with filters to exclude what they have already seen.
Data from GitHub and our catalog. Every listing is hand-approved. List prices are what the vendor publishes, not a quote, and self-hosting has its own running costs.
