What is this?
A lightweight ~20 MB cross-platform database client covering 90+ databases — MySQL, PostgreSQL, SQLite, Redis, MongoDB, DuckDB, ClickHouse, SQL Server and more — with a built-in AI assistant and an MCP server.
A lightweight ~20 MB cross-platform database client covering 90+ databases — MySQL, PostgreSQL, SQLite, Redis, MongoDB, DuckDB, ClickHouse, SQL Server and more — with a built-in AI assistant and an MCP server.Is it a desktop app, or does it run in a browser too?It's a native desktop app on macOS, Windows, and Linux, plus a web version for browser-only environments. You can also self-host it via Docker so a team shares connections, and a CLI is included.
It's a native desktop app on macOS, Windows, and Linux, plus a web version for browser-only environments. You can also self-host it via Docker so a team shares connections, and a CLI is included.How do I get it running?Download the single small binary, install it, and connect to a database — that's the whole process. There's no runtime to set up and no build step for end users.
Download the single small binary, install it, and connect to a database — that's the whole process. There's no runtime to set up and no build step for end users.Do I need to install Java or anything else first?No. There is no Java JRE, no Python venv, and no bundled Chromium — unlike DBeaver, which requires Java. Node 22.13+ and pnpm are only needed if you build it from source.
No. There is no Java JRE, no Python venv, and no bundled Chromium — unlike DBeaver, which requires Java. Node 22.13+ and pnpm are only needed if you build it from source.How does it compare to DBeaver or TablePlus?DBeaver needs Java and TablePlus is freemium, while DBX is a single ~20 MB binary that runs with nothing extra. If you're switching over, it can import connection profiles from DBeaver or Navicat.
DBeaver needs Java and TablePlus is freemium, while DBX is a single ~20 MB binary that runs with nothing extra. If you're switching over, it can import connection profiles from DBeaver or Navicat.Do the AI features send my queries to the cloud?Only to the AI provider you pick — Claude, OpenAI, or any OpenAI-compatible endpoint. You can point it at local models via Ollama to keep everything on your machine, and config export/import is encrypted.
Only to the AI provider you pick — Claude, OpenAI, or any OpenAI-compatible endpoint. You can point it at local models via Ollama to keep everything on your machine, and config export/import is encrypted.Which databases can it connect to?90+ out of the box: MySQL, PostgreSQL, SQLite, Redis, MongoDB, DuckDB, ClickHouse, SQL Server, Oracle, Elasticsearch, MariaDB, TiDB, CockroachDB, plus Chinese databases like Dameng, OceanBase, and KingbaseES. Agent-based profiles extend it to Snowflake, Trino, Hive, BigQuery, and custom JDBC connections, with specialized browsers for Redis and MongoDB.
90+ out of the box: MySQL, PostgreSQL, SQLite, Redis, MongoDB, DuckDB, ClickHouse, SQL Server, Oracle, Elasticsearch, MariaDB, TiDB, CockroachDB, plus Chinese databases like Dameng, OceanBase, and KingbaseES. Agent-based profiles extend it to Snowflake, Trino, Hive, BigQuery, and custom JDBC connections, with specialized browsers for Redis and MongoDB.What can the built-in AI assistant actually do?Describe what you want in plain language and it writes the SQL; it can also explain, optimize, and fix queries, and generated SQL goes through built-in safety checks before it runs. It works with Claude, OpenAI, local models, or any OpenAI-compatible endpoint.
Describe what you want in plain language and it writes the SQL; it can also explain, optimize, and fix queries, and generated SQL goes through built-in safety checks before it runs. It works with Claude, OpenAI, local models, or any OpenAI-compatible endpoint.What is the MCP server for?It lets AI coding agents like Claude Code, Cursor, and Windsurf query your databases directly, reusing the connections you already configured in DBX — one config works everywhere.
It lets AI coding agents like Claude Code, Cursor, and Windsurf query your databases directly, reusing the connections you already configured in DBX — one config works everywhere.Can I edit data inline, and what formats can I export to?Yes — the data grid supports inline editing with an SQL preview before saving, plus filters, sorting, and full-text search. Results export or copy as CSV, JSON, Markdown, XLSX, or INSERT statements; you can import CSV and Excel, drag-and-drop Parquet/CSV/JSON files for instant preview via DuckDB, and destructive operations ask for confirmation.
Yes — the data grid supports inline editing with an SQL preview before saving, plus filters, sorting, and full-text search. Results export or copy as CSV, JSON, Markdown, XLSX, or INSERT statements; you can import CSV and Excel, drag-and-drop Parquet/CSV/JSON files for instant preview via DuckDB, and destructive operations ask for confirmation.What are similar projects?Chat2DB, TablePro, tusk, tabularis, GoNavi, navop, beekeeper-studio, mcp-toolbox
Chat2DB, TablePro, tusk, tabularis, GoNavi, navop, beekeeper-studio, mcp-toolboxWhich topics is it in?database-management, database-gui, oracle, sql-server, duckdb, database-client, clickhouse, mariadb, elasticsearch, mongodb, tauri, ssh
database-management, database-gui, oracle, sql-server, duckdb, database-client, clickhouse, mariadb, elasticsearch, mongodb, tauri, ssh