Cohere – Use Cohere embeddings with Qdrant (blogpost)
DocArray – Use Qdrant as a document store in DocArray
Haystack – Use Qdrant as a document store with Haystack (blogpost)
LangChain – Use Qdrant as a memory backend for LangChain (blogpost)
LlamaIndex – Use Qdrant as a vector store with LlamaIndex
OpenAI – ChatGPT retrieval plugin – Use Qdrant as a memory back end for ChatGPT
Microsoft Semantic Kernel – Use Qdrant as persistent memory with Semantic Kernel
Vector search benchmarks
Qdrant ran a bunch of benchmarks that compared Qdrant with open-source rivals Weaviate, Elasticsearch, Redis, and Milvus, each running on a single node with 100 search threads. You can explore the results online. While I rarely trust vendor benchmarks of other vendors’ products, Qdrant has gone to some trouble to make these reproducible.
Qdrant Cloud
Qdrant Cloud runs fully managed clusters on multiple regions of Amazon Web Services, Google Cloud Platform, and Microsoft Azure. It offers central cluster management, horizontal and vertical scaling, high availability, auto-healing, central monitoring, log management and alerting, backup & disaster recovery, zero-downtime upgrades, and unlimited users. The screenshots below illustrate many of Qdrant Cloud’s capabilities.
Creating a Qdrant cluster is a simple matter of filling out a web form. This is a paid cluster, which has more options than a free cluster. Google Cloud Platform is currently the least expensive of the three supported cloud providers.
IDG
The collections I currently have loaded into my free Qdrant Cloud cluster came from running demo or tutorial code, or from loading sample data sets.
IDG
The Qdrant Cloud console allows you to run requests against your clusters interactively. Notice the Run|Beautify|Docs links above the code. The results of the most recent request are displayed at the right. The buttons at the lower right bring up additional commands.
IDG
Here we are viewing the points in the mid-styles collection, which include the links to images in their payloads. Note that the vectors are not displayed. You wouldn’t want to look at 512-length vectors in the web interface, but if you need to see them you can copy them and paste them into an editor. You can perform a vector similarity search from any point by clicking the “Find Similar” button at the right.
IDG
The Qdrant Cloud interactive tutorial provides a console with the code already loaded. You can click on the Run button for each block to proceed through the steps. The result displayed at the right is from the simple Post vector search query at the left.
IDG
Qdrant allows filtering by payload as well as vector search, and can do both simultaneously. Filter keywords include “must,” which is a Boolean AND operator; “should,” which is a Boolean OR operator, and “must not,” which is a negation operator as explained at the lower left. The results at the right are from a “should” filter.
IDG
You can view the Qdrant Cloud cluster status from the cluster details. This is my free cluster, which as you can see is healthy and is only using a fraction of its capacity.
IDG
You can also view the requests to the cluster over time, and see their latency, which is very low for these requests because I’m just playing around interactively, not running an application with a heavy load.
IDG
You can create a visualization of the points in a collection, filter them, limit the number of points displayed (important for large collections), and color the points based on the payload values. The visualization uses a t-SNE dimensionality reduction algorithm to display the high-dimensional vectors.
IDG
If your Qdrant Cloud cluster isn’t a good fit for your application, you can scale it horizontally and/or vertically. If you have more than one node in your cluster, you can scale it without incurring down time.
IDG
Qdrant hybrid and private clouds
You aren’t required to run Qdrant exclusively as SaaS on public clouds to use its own, very convenient managed cloud interface. You can also run Qdrant in Kubernetes clusters under your own control, either on-premises or in cloud instances, and connect those to the management interface running in the public cloud. Qdrant calls this hybrid cloud. In addition, you can run Qdrant in private cloud configurations, again either on-prem or in cloud instances, with the cluster management running in your own infrastructure, in the public cloud, on premises at the edge, and even fully air-gapped.