SQL-friendly, Cost-effective Vector Database
Trusted SQL/Integrated Vector Database
Data Foundation to Maximize AI Application Performance
Store and manage structured and vector data in one system with powerful and familiar SQL.
Numeric, date/time, text, geospatial, JSON, vector, time series all supported.
Combine structured and vector querying for high accuracy RAG.
Translate text into flexible and precise SQL & vector joint queries.
Perform vectorization and reranking AI functions using SQL.
Store AI agent logs and traces in database for observability and continuous improvement.
High-Performance and Cost-Efficient
A single c1x1 pod supports up to 10 million 768D vectors, while the s1x1 pod achieves over 150 QPS with 5 million vectors.
Ingest & build index for 5 million points in under 30 minutes for faster serving.
Manage multiple tables with different vector indexes in one pod, reducing complexity and cost.
Support for Parquet, CSV/TSV and compressed tar files facilitates seamless data transfer with a single SQL command.
What Can You Build with MyScale
Comparison
- SQL and Vector Support
- Metadata support
- High accuracy filtered search
- High performance filtered search
- Hybrid search support
- Cost-performance The metric reference is calculated as the monthly cost per 100 QPS. Lower costs represent higher cost-effectiveness.
- Cost-performance for filtered-search The metric reflects the monthly cost for achieving 100 QPS with metadata filtering. Lower costs represent higher cost-effectiveness.
Security and Compliance
Access the benefits of SaaS with confidence in security. MyScale provides end-to-end protection, empowering your organization to efficiently and securely share data with others, ensuring peace of mind throughout the process.