DBX: A 15MB "Powerhouse" That Packs 50+ Database Types

Follow my public account: [Programming Morning Blossom Evening Harvest] for first-release content.

01 Introduction

Why another database client? The database client space is as competitive as frontend frameworks. The three giants — Navicat, DataGrip, DBeaver — loom large, with a swarm of self-proclaimed "disruptors" trailing behind.

We previously covered gonavi. Today, let's look at another product: DBX.

02 Overview

DBX is a lightweight, multi-database management client that packs nearly every capability needed for daily database operations into one product:

• Connection management
• SQL editing and querying
• Data browsing and editing
• Database schema tools
• AI assistant
• Self-hosted web access

It offers two deployment modes:

Mode	Use Case
Desktop App	Local development, personal use
Docker Self-Hosted	Team sharing, browser access, intranet deployment

And both modes come from the same project, meaning the experience you get locally is identical to what you deploy for your team. This is very rare among open-source tools.

Official site: https://dbxio.com/cn

GitHub: https://github.com/t8y2/dbx

03 Core Highlights

3.1 Extremely Lightweight

An installer of about 15MB. DBX is built natively in Rust and does not depend on a JDBC runtime.

This means:

• The installer is ridiculously small (about 15MB, compared to Navicat's 200MB+ and DataGrip's hundreds of MB)
• Startup speed is noticeably fast
• Memory usage is far lower than JVM-based clients
• No more waiting for JVM startup before you can get to work

3.2 Supports 50+ Database Engines

DBX isn't picky — from mainstream relational databases to time-series databases, search engines, and domestic Chinese databases, it covers almost everything:

International Mainstream

• MySQL, PostgreSQL, SQLite, SQL Server, Oracle, DB2
• MongoDB, Redis, Elasticsearch, Neo4j, Cassandra
• ClickHouse, Snowflake, BigQuery, Databricks, Redshift
• Trino, Hive, Databend, DuckDB

Domestic Chinese Databases

• Dameng, GaussDB, openGauss, KingBase
• HighGo, TiDB, OceanBase, SelectDB
• TDengine, KWDB, Vastbase, GoldenDB, YashanDB
• GBase, XuguDB, SunDB

Time-Series / IoT

• InfluxDB, QuestDB, IoTDB

Others

• etcd, IRIS, JDBC generic connection...

Coverage is on par with DBeaver, but startup speed and size leave it in the dust.

3.3 AI Support

After configuring an AI provider, you can generate SQL, explain queries, optimize statements, and get help fixing errors within DBX.

As shown, DBX supports numerous large model providers. It also supports MCP.

04 Quick Look

Clean and refreshing interface:

Viewing data:

DDL:

Query:

AI Chat:

Permission Management:

05 Summary

DBX isn't a "disruptor" — it's a "pragmatist." In a space bloated by feature-heavy tools, it manages 50+ database types with a 15MB footprint, cleanly and clearly.

If you're tired of Navicat's bloat, DataGrip's memory hunger, or DBeaver's slow startup, DBX is worth a shot.