How to Choose a Vector Database: Qdrant Cloud vs. Zilliz Cloud
Compare Qdrant Cloud and Zilliz Cloud (fully managed Milvus) in this in-depth benchmark, cost, and features comparison.
Read the entire series
- Introducing an Open Source Vector Database Benchmark Tool for Choosing the Ideal Vector Database for Your Project
- How to Choose A Vector Database: Elastic Cloud vs. Zilliz Cloud
- How to Choose a Vector Database: Qdrant Cloud vs. Zilliz Cloud
- How to Choose A Vector Database: Weaviate Cloud vs. Zilliz Cloud
- Benchmarking Vector Database Performance: Techniques and Insights
- Developing Custom Applications with Vector Databases
This Qdrant comparison was last updated on October 13, 2023. To provide you with the latest findings, this blog will be regularly updated with the latest benchmark figures.
Not all vector databases are made equal: unveiling the distinction
In the fast-paced world of vector databases, purpose-built solutions are increasingly becoming the norm, overshadowing traditional databases with hastily added vector search plugins. Qdrant, although newer to the scene, has quickly gained attention for its ease of use and developer-friendly documentation. Built entirely in Rust, it offers APIs accessible via Rust, Python, and Golang clients, catering to the most popular languages for backend developers today. However, being a relatively newer tool than its competitors, Qdrant has been playing catch-up with alternatives in terms of lots of key features such as UI and querying. Zilliz/Milvus, on the other hand, distinguishes itself as a mature, extremely large-scale, high-performance, low-latency vector database, optimized for the most demanding applications.
While Qdrant and Zilliz/Milvus are purpose-built for vector data, they serve different market needs. Qdrant appeals to developers who prioritize modern technology and seek to minimize infrastructure maintenance. Zilliz/Milvus, on the other hand, is engineered for extreme scale, high performance, and low latency, making it a robust and innovative solution for applications that require top-tier performance and scalability.
We will venture beyond these distinctions within this Qdrant Cloud vs. Zilliz Cloud comparison, delving into several other comparison areas. We'll delve into benchmarks to offer a performance perspective and perform an in-depth Qdrant Cloud feature analysis.
Benchmarking Qdrant Cloud and Zilliz Cloud
At Zilliz, developers and architects alike ask us, "How does Zilliz measure up to Qdrant Cloud for vector embedding workloads?" This curiosity might spring from various motivations, all stemming from building semantic similarity search for capabilities like product recommenders or reverse image search, and lately, when building LLM applications with Retrieval Augmented Generation (RAG).
Over the past few weeks, we undertook a comprehensive analysis, delving into the performance and attributes of Zilliz Cloud vs. Qdrant Cloud regarding vector embedding workloads. With our open-source benchmark tool, VectorDBBench, our exploration focused intently on critical aspects such as queries per second (QPS), queries per dollar (QP$), and latency.
Large-size datasets tested (≥5M vectors)
Dataset 1: 10,000,000 vectors with 768 dimensions
Dataset 2: 5,000,000 vectors with 1,536 dimensions
Products tested (with similar capabilities)
Zilliz Cloud (8cu-perf): Zilliz Cloud with eight performance-optimized compute units
Zilliz Cloud (2cu-cap): Zilliz Cloud with two capacity-optimized compute units
Qdrant Cloud (4c16g-5node): Qdrant Cloud with four CPUs and 16G of RAM using five machines
For more information about Zilliz Cloud’s compute units, see Zilliz blog introducing Zilliz Cloud CU type and size.
Results: QPS
When processing 10 million vectors with 768 dimensions, Zilliz Cloud options outperform Qdrant Cloud by over 7x and 1x, respectively, in QPS.
When processing 5 million vectors with 1536 dimensions, Zilliz Cloud(8cu-perf) outperforms Qdrant Cloud by almost 8x in QPS, while Zilliz Cloud (2cu-cap) shows lower performance than Qdrant Cloud.
Results: QP$
When processing 10 million vectors with 768 dimensions, Zilliz Cloud options outperform Qdrant Cloud by over 8x and 5x respectively in QP$.
When processing 5 million vectors with 1536 dimensions, Zilliz Cloud options outperform Qdrant Cloud by 7x and 3x respectively in QP$.
Results: Latency
When processing 10 million vectors with 768 dimensions, Zilliz Cloud options outperform Qdrant Cloud by over 12x and 3x respectively in latency.
When processing 5 million vectors with 1536 dimensions, Zilliz Cloud options outperform Qdrant Cloud by over 8x and 1x respectively in latency.
Medium-size datasets tested (< 5M vectors)
- Dataset 3: 1,000,000 vectors with 768 dimensions
- Dataset 4: 500,000 vectors with 1,536 dimensions
Products tested (with similar capabilities)
Zilliz Cloud (1cu-perf): Zilliz Cloud with one performance-optimized compute unit
Zilliz Cloud (1cu-cap): Zilliz Cloud with one capacity-optimized compute unit
Qdrant Cloud (4c16g-1node): Qdrant Cloud with four CPUs and 16G of RAM using one machine
For more information about Zilliz Cloud’s compute units, see Zilliz blog introducing Zilliz Cloud CU type and size.
Results: QPS
When processing 1 million vectors with 768 dimensions, Zilliz Cloud options outperform Qdrant Cloud by over 2x and 1x respectively in QPS.
When processing 500,000 vectors with 1,536 dimensions, Zilliz Cloud(1cu-perf) outperforms Qdrant Cloud by almost 2x in QPS, while Zilliz Cloud (1cu-cap) shows lower performance than Qdrant Cloud.
Results: QP$
When processing 1 million vectors with 768 dimensions, Zilliz Cloud options outperform Qdrant Cloud by over 4x and 2x respectively in QP$.
When processing 500,000 vectors with 1,536 dimensions, Zilliz Cloud options outperform Qdrant Cloud by over 3x and 1x respectively in QP$.
Results: Latency
When processing 1 million vectors with 768 dimensions, both Zilliz Cloud potions outperform Qdrant Cloud by more than 2x in latency.
When processing 500,000 vectors with 1,536 dimensions, Zilliz Cloud potions outperform Qdrant Cloud by more than 127x and 27x respectively in latency.
Results: Comprehensive Scores
Note: This is a 1-100 score generated by VectorDBBench based on each system's performance in different cases according to a specific rule. A higher score denotes better performance.
Note: This is a >1 score by VectorDBBench based on each system's performance in different cases according to a specific rule. A lower score denotes better performance.
We conducted these benchmarks using the open-source VectorDB Bench tool on GitHub, where you’ll also find the updated leaderboard. VectorDBBench provides unbiased vector database benchmark results for mainstream vector databases and cloud services. Designed for ease of use, VectorDB Bench makes finding the optimal choice among a plethora of vector database cloud services and open-source vector databases a breeze.
Qdrant Cloud features comparison
Performance is the biggest challenge with vector databases as the number of unstructured data elements stored in a vector database grows into hundreds of millions or billions, and horizontal scaling across multiple nodes becomes paramount. Furthermore, differences in insert rate, query rate, and underlying hardware may result in different application needs, making overall system tunability a mandatory feature for vector databases.
What’s your vector database for?
A vector database is a fully managed solution for storing, indexing and searching across a massive dataset of unstructured data that leverages the power of embeddings from machine learning models. A vector database should have the following features:
Scalability and tunability
Multi-tenancy and data isolation
A complete suite of APIs
An intuitive user interface/administrative console
Scalability
| Zilliz Cloud | Qdrant Cloud | |
|---|---|---|
| Billion-scale vector support | √ | x |
| Horizontal scaling | √ | √ |
| Separation of storage and compute | √ | x |
| Dynamic segment placement vs. static data sharding | Dynamic segment placement | Static sharding |
| Separation of query and insertions | Yes. At the component level (which provides more fine-grained scalability). | No. Only scale at the server level. |
Functionality
| Zilliz Cloud | Qdrant Cloud | |
|---|---|---|
| Role-based Access Control (RBAC) | √ | No. Authentication only |
| Disk Index support | √ | √ |
| Hybrid Search (ie, Scalar filtering) | Yes with Scalar filtering | Yes (combine vector and traditional indices) |
| Partition-key support | √ | x |
Purpose-built
| Features | Zilliz Cloud | Qdrant Cloud |
|---|---|---|
| Purpose-built for vectors | √ | √ |
| Multiple consistency choices | √ | x |
| Support for both stream and batch of vector data | √ | x |
| Binary vector support | x | x |
| Multi-language SDK | Python, Java, Go, C++, Node.js, Ruby | Python, Go, Rust |
For more details, visit our Milvus vs Qdrant comparison page.
What’s next?
Steffi LiSteffi is the Director of Product Marketing at Zilliz, with prior experience leading the GTM strategies for open-source data technologies such as Apache Kafka and Apache Airflow. Passionate about travel, yoga, and non-fiction reading, she brings a balanced and enriched worldview to her professional pursuits.
- Not all vector databases are made equal: unveiling the distinction
- Benchmarking Qdrant Cloud and Zilliz Cloud
- Qdrant Cloud features comparison
- What’s next?
Content
Start Free, Scale Easily
Try the fully-managed vector database built for your GenAI applications.
Try Zilliz Cloud for Free