DBeaver: Universal Database Tool and SQL Client

What DBeaver is

DBeaver is a free, multi-platform database tool for developers, SQL programmers, database administrators and analysts. It is a desktop application that puts a schema editor, SQL editor, data editor, ER diagrams, data export, import and migration, SQL execution plans, administration tools, dashboards and a spatial data viewer behind one interface, and it connects to almost anything: the Community edition ships with more than 100 database drivers out of the box and supports any database that has a JDBC or ODBC driver.

The problem it solves is familiar to anyone who works across more than one database. Each engine tends to come with its own client, its own shortcuts and its own export formats, and switching between them every day adds friction. DBeaver gives a single, consistent workspace for PostgreSQL in the morning, SQLite in the afternoon and a Snowflake warehouse in the evening, with SSH tunnelling and proxy support for reaching databases that are not exposed directly.

The project has been on GitHub since October 2015, is written mostly in Java and is licensed under Apache-2.0. It also integrates AI tools for working with data, SQL and database structure.

How it works

DBeaver is built on well-established Java desktop and plugin foundations. The README describes the stack:

• OSGi for plugins and dependency management. The Community version alone consists of more than 130 plugins.
• Eclipse RCP as the rich-client platform for the user interface.
• JDBC as the basic database connectivity API.
• JSQLParser and ANTLR 4 for SQL grammar and semantic parsing, which power features such as completion and analysis in the SQL editor.
• A wide set of open-source libraries for networking and extra features, including SSHJ, Apache POI, JFreeChart, JTS and Apache JEXL.

Alongside the Java code it uses native, OS-specific components for the desktop UI, high-performance database drivers and networking. A key design choice is the separation of model plugins from desktop UI plugins: the same set of back-end plugins is reused by CloudBeaver, the team's web-based database manager, so the connection and metadata logic is shared between desktop and web. Third-party dependencies come from P2 repositories, with extra Maven dependencies served from a dedicated DBeaver P2 repository.

DBeaver needs Java to run, but every distribution bundles OpenJDK 25, so there is nothing separate to install. The bundled runtime lives in the jre directory of the installation folder, and replacing that directory changes the JDK the application uses.

Key features

• SQL editor: write and run queries with parsing-backed assistance and view execution plans.
• Data editor: browse and edit table data directly in a grid.
• Schema editor and ER diagrams: inspect and change database structure and visualise relationships.
• Data transfer: export, import and migrate data between sources.
• Administration tools and dashboards: monitor and manage database servers from the same client.
• Spatial data viewer: inspect spatial (GIS) data alongside regular query results; JTS is among the libraries the project builds on.
• Connectivity options: SSH tunnelling, proxies and a custom database drivers editor for engines that are not supported out of the box.
• AI chat: an AI Chat view that can generate, analyse and optimise SQL and work with database structure. The README says it supplies the model with context such as the schema and SQL dialect, and uses context-dependent tools to keep token consumption low.
• Broad driver coverage: the Community edition list runs from PostgreSQL, MySQL, MariaDB, SQL Server, Oracle and SQLite to ClickHouse, DuckDB, CockroachDB, Snowflake, BigQuery, Databricks, Trino, TiDB and Yugabyte.

Getting started

Prebuilt binaries are available from the official website and from GitHub releases, and an Early Access build is published daily. Running DBeaver is a matter of launching the installer and clicking the app icon, or unzipping an archive and running dbeaver from the command line.

To build from source, the project wiki asks for JDK 25 or newer, Git and internet access for downloading dependencies. Clone the repository:

git clone https://github.com/dbeaver/dbeaver.git

Then run the build script for your platform. On Linux or macOS:

./dbeaver/tools/build.sh

On Windows:

.\dbeaver\tools\build.cmd

When the build completes, the binaries are in the product/community/target/products directory.

Use cases

• Polyglot backend teams: one client for the transactional database, the analytics warehouse and the embedded SQLite files used in tests.
• Database administration: review execution plans, manage objects and keep an eye on server dashboards without a vendor-specific console.
• Data migration: move data between engines with the built-in export, import and migration tooling.
• Schema reviews: generate ER diagrams to understand an unfamiliar database before changing it.
• Remote databases behind bastions: connect through SSH tunnels or proxies to databases that are not exposed to the internet.
• SQL help for non-specialists: use the AI Chat view to draft or explain queries against a schema the model can see.

How it compares

The DBeaver team positions its own products side by side. CloudBeaver is a web-based database manager built on the same platform that runs as a server (for example in Docker) and serves a single-page web interface, which suits teams that want browser access instead of a desktop install. dbvr is a command-line database management tool aimed at CI/CD pipelines and automation. Beyond the DBeaver family, the usual alternatives are engine-specific clients that come with a particular database, which tend to go deeper on one engine, while DBeaver trades some of that depth for breadth across many engines in a single tool.

Things to know before adopting

• Community versus PRO: the Apache-2.0 Community edition covers relational and many analytical engines. NoSQL support (MongoDB, Redis, Cassandra, DynamoDB, Neo4j and others), files-as-databases for formats like CSV, JSON, Parquet and XLSX, and a federated multi-source database based on Apache Calcite are in the commercial editions.
• AI providers: the Community edition supports OpenAI (including other LLMs through a custom endpoint) and Copilot. Native support for Anthropic, Grok, Azure, Bedrock, Gemini and Ollama providers is in the Pro versions.
• Desktop footprint: as an Eclipse RCP application with a bundled JDK, it is a full desktop install rather than a lightweight utility.
• Default branch: development happens on the devel branch, so check release tags rather than the branch head when you want a stable build from source.
• Support model: community support runs through GitHub issues and discussions, where tickets can be voted up; official online support and priority bug fixes come with the commercial licences.

Project activity

As of October 2026 the repository has about 52,000 stars. It was created in October 2015, is written in Java and is licensed under Apache-2.0, with pushes as recent as mid-October 2026 and an Early Access build published every day. The team labels tickets as good first issues and help wanted for new contributors. The source is at github.com/dbeaver/dbeaver, the product site at dbeaver.io and the full documentation at dbeaver.com/docs/dbeaver.