README.md
May 31, 2026 · View on GitHub
DbPaw
Faster SQL editing and data exploration — cross-platform, ultra-lightweight, with optional AI assistance.
English | 简体中文
DbPaw helps you connect to PostgreSQL, MySQL, MariaDB (MySQL-compatible), TiDB (MySQL-compatible), SQLite, SQL Server, ClickHouse (preview), DuckDB, StarRocks, Doris, Oracle, and Redis, write and run SQL efficiently, and inspect data in a clean desktop UI.
✅ What You Can Do
- Connect to PostgreSQL, MySQL, MariaDB (MySQL-compatible), TiDB (MySQL-compatible), SQLite, SQL Server, ClickHouse (preview, currently read-only), DuckDB, StarRocks, Doris, Oracle, and Redis (Standalone / Cluster / Sentinel)
- Write and run SQL with syntax highlighting, auto-completion, and one-click formatting
- Browse query results in a data grid with filtering, sorting, and pagination
- Export table data and query results as CSV, JSON, or SQL (DDL only / DML only / DDL+DML); choose current page, filtered rows, or the full table
- Export an entire database as a SQL file (schema + data)
- Import
.sqlfiles into MySQL/MariaDB/TiDB/PostgreSQL/SQLite/DuckDB/SQL Server with all-or-nothing rollback - Visually create and alter table structures without writing DDL by hand
- Inspect table structure — columns, types, primary keys, indexes — and view the live DDL
- Track every query in the SQL execution log with timing and status
- Save and reuse frequently used SQL scripts with Saved Queries
- Use the AI sidebar to draft SQL and explain queries (optional)
- MCP Server — full MCP 2025-03-26 protocol support with stdio + HTTP dual-mode transport, letting AI assistants (Claude, Cursor) directly query your databases (docs)
- Access remote databases through SSH tunneling
- Browse and manage Redis data: keys, Strings, Hashes, Lists, Sets, Sorted Sets, Streams, and JSON — with Cluster and Sentinel support
🖼️ Screenshots


| Connection | SQL Editor |
|---|---|
![]() | ![]() |
| Data Grid | AI Assistant |
|---|---|
![]() | ![]() |
✨ Features
- Tiny footprint: installer ≈10 MB, on-disk ≈80 MB, and very low idle memory (much lighter than Electron-based tools).
- Truly modern: goodbye to DBeaver-style “cockpit” UIs—we cut the 99% you’ll never use, focus on common workflows, and make every action smoother and more intuitive.
- Cross-platform: runs on macOS, Windows, and Linux (no more one app at work and another at home).
- Database compatibility: currently supports MySQL, MariaDB (MySQL-compatible), PostgreSQL, ClickHouse, TiDB, SQL Server, SQLite, DuckDB, StarRocks, Doris, Oracle, and Redis (Standalone / Cluster / Sentinel) — actively expanding.
- Rich data transfer: export to CSV / JSON / SQL (DDL, DML, or both) with flexible scope; import SQL files with transactional rollback; dump entire databases as SQL.
- Schema explorer: browse tables, view structure and DDL, and create or alter tables through a visual UI—no hand-crafted DDL required.
- Looks great: lots of themes (dark/light and a range of high/low saturation styles).
- Built-in AI assistance (experimental): summarize SQL, explain schemas, and analyze slow queries with AI (security under active refinement; local/optional cloud modes planned).
- Completely free: no login, no payments, no memberships, no ads.
📥 Installation
Go to the Releases page to download the latest version for your operating system.
macOS Users
- Download
DbPawfor macOS from Releases. - Move
DbPaw.appto your/Applicationsfolder. - Open the app.
If macOS blocks the app with an "Unidentified Developer" warning:
- Open System Settings → Privacy & Security.
- Scroll to the Security section and find the message about
DbPawbeing blocked. - Click Open Anyway, then confirm Open.
If you encounter a "DbPaw is damaged" warning (Gatekeeper quarantine):
- Move
DbPaw.appto your/Applicationsfolder. - Open Terminal and run the following command:
sudo xattr -d com.apple.quarantine /Applications/DbPaw.app - You can now open the app normally.
Note: This is required because the app is not yet notarized by Apple.
Windows Users
- Download the installer or portable build from Releases.
- Run the installer / executable.
If Windows shows a security warning such as "Windows protected your PC" (SmartScreen):
- Click More info.
- Click Run anyway.
If your device is managed by an organization, you may need your IT admin to allow the app.
🔐 Security & Privacy
- DbPaw is a local desktop app. Your database connections run from your machine to your database.
- AI features are optional. When enabled, DbPaw sends your prompt, recent chat context, and (optionally) a schema overview (tables/columns/types) to the AI provider you configured.
- AI conversations are stored locally. AI provider API keys are stored encrypted on disk.
- No built-in telemetry/analytics SDK is included in the desktop app.
🛠️ Development
- Development guide: docs/en/Development/DEVELOPMENT.md
- Contributing guide: docs/en/Community/CONTRIBUTING.md
- MCP Server guide: docs/mcp.md
🏗️ Tech Stack
- Core: Tauri v2 (Rust)
- Frontend: React 19, TypeScript
- Styling: TailwindCSS v4, Shadcn/UI
- State Management: React Hooks & Context
- Editor: Monaco Editor / CodeMirror
📄 License
This project is licensed under the Apache License 2.0 - see the LICENSE file for details.
Oracle Instant Client
This software includes Oracle Instant Client, which is subject to the Oracle Technology Network License Agreement for Oracle Instant Client. Oracle Instant Client is free to use for connecting to Oracle Database.
❤️ Thanks
Thanks for trying DbPaw. If you find it useful, please consider giving this repository a star — it helps a lot!



