ClickHouse: Open-Source Column-Oriented OLAP Database

ClickHouse is described as an open-source columnar database management system designed for real-time analytical queries on massive datasets, using SQL. Data is stored in columns, with values from the same column stored together, which the company states makes column-oriented databases better suited to OLAP scenarios and at least 100x faster in processing most queries. The engine uses all available system resources to process each analytical query, scaling both horizontally and vertically across hundreds of cores and petabytes of storage. Monthly releases add features, performance improvements and bug fixes, and are discussed in community release webinars.

• Open-source, column-oriented SQL database
• Single binary for analytics, warehousing, observability and GenAI
• Parallelised query execution and column-oriented compression
• Monthly releases with community webinars

ClickHouse can be installed for macOS, Linux and FreeBSD with a single shell command, and install options also exist for Windows and Docker. Those who prefer not to install anything can try the online playground. For local files, clickhouse-local provides a portable tool to query, convert and transform data on a local machine, and chDB makes ClickHouse available as an in-process OLAP SQL engine for Python.

• Install command: curl https://clickhouse.com/ | sh
• Platforms: macOS, Linux, FreeBSD, Windows, Docker
• clickhouse-local for querying local files with SQL
• chDB for in-process use inside Python code
• Online playground for trying without installing

ClickHouse groups its workloads into four main use cases: real-time analytics, observability, data warehousing, and machine learning and GenAI. Real-time analytics covers user-facing dashboards and applications, continuous ingest from streaming sources and high query concurrency. Data warehousing covers analysing data in more than 70 file formats, integrating with visualisation and transformation tools, and paying only for the compute and compressed storage used. The ML and GenAI use case covers aggregations for data preparation, vector search with linear and approximate techniques, user-defined functions for inference, and agent-facing analytics via the ClickHouse Cloud console or a remote MCP server.

• Real-time analytics: dashboards, apps, streaming ingest, high concurrency
• Data warehousing: 70+ file formats from Parquet to JSON, CSV and TSV
• Observability: logs, metrics and traces via ClickStack
• ML and GenAI: vector search, aggregations, pre-built models, MCP server
• Integrations with Grafana, Tableau, Superset, Metabase and dbt

ClickStack is ClickHouse's open-source observability stack for OpenTelemetry, unifying logs, metrics, traces, session replays and errors. It combines ClickHouse as the columnar store, HyperDX as the unified UI for search, dashboards and alerts, and OpenTelemetry for standardised data collection. The UI supports Lucene-style search alongside full SQL access, with more than 100 built-in functions, event deltas for anomaly detection and event patterns for root cause analysis. A Managed ClickStack option runs on ClickHouse Cloud, where separation of storage and compute supports long-term retention and independent scaling of ingestion and query resources.

• Components: ClickHouse, HyperDX, OpenTelemetry
• Optimised OpenTelemetry schemas shipped by default
• Native JSON support for evolving, semi-structured data
• Managed ClickStack available on ClickHouse Cloud
• Compute-compute separation for ingestion and querying

ClickHouse publishes industry pages for financial services and gaming. In financial services it addresses capital markets, banking and payments, financial crime, and digital assets and crypto, covering quant backtests, real-time P&L, fraud decisioning, AML monitoring and KYC pattern scoring. For gaming it targets high-throughput analytics and telemetry, including player behaviour analysis, monetisation and ad decisions, and game performance monitoring for gaming and wagering platforms. Both pages cite native OpenTelemetry support, streaming ingestion from sources such as Kafka, Kinesis and Pub/Sub, and controls including CMEK, BYOC support, fine-grained access controls and GDPR-compliant TTLs and deletes.

• Financial services: capital markets, banking and payments, financial crime, digital assets
• Gaming: in-game event telemetry, leaderboards, ad tech, crash analytics
• Agentic analytics via ClickHouse MCP and LibreChat
• Compliance and deployment controls for regulated institutions

ClickHouse documents migration routes from several other analytical systems, each framed around performance, cost and architectural simplification. Migrating from Elastic is positioned around faster aggregations and a smaller storage footprint; Apache Druid around performance and infrastructure costs; Snowflake around cost and latency without tier-gated features; BigQuery around avoiding per-query pricing and cloud lock-in; Redshift around faster cold queries and broader integrations; and Apache Pinot around richer aggregation functions and full SQL compliance. Supporting comparison material includes benchmark write-ups against Snowflake, BigQuery and Redshift.

• Elasticsearch
• Apache Druid
• Snowflake
• BigQuery
• Redshift
• Apache Pinot

House Mates is the ClickHouse partner community, open to organisations that sell, integrate or build on the database. It offers partner incentives with margins and deal registration, joint go-to-market activity with listing in a partner directory, and enablement including sales kits, sandbox environments, certifications, instructor-led workshops and on-demand learning paths. The programme lists five partner types, and partners can apply to join, search the partner directory or access the partner portal. The page states 4000+ ClickHouse Cloud customers and 49.9k GitHub stars.

• Cloud Partners — hyperscalers and cloud marketplaces including AWS, Microsoft Azure and Google Cloud
• Consulting & Services — SIs and consultancies implementing and migrating deployments
• Technology Partners — connectors, drivers and integrations
• Reseller Partners — VARs, distributors and managed service providers
• Powered by ClickHouse — ISVs and SaaS companies embedding ClickHouse

The website offers a free ClickHouse Cloud trial as the main entry point, alongside a 'Contact sales' route and, on industry pages, a form to get in touch with a ClickHouse expert about a specific use case. Open-source users can install the binary directly or start with the playground. Managed ClickStack is also available with a free trial on ClickHouse Cloud.

• Free ClickHouse Cloud trial
• Contact sales enquiry route
• Expert contact forms on industry pages
• Open-source download and playground