DB-GPT
Open-source AI assistant that writes SQL and code to analyze your databases and files.
- Stars
- 20.2k
- Forks
- 3k
- Licence
- MIT
- Language
- Python
- Last update
- 4 days ago
- Type
- App
What it is
DB-GPT is an open-source AI data assistant. You connect it to databases, CSV or Excel files, warehouses, or knowledge bases, then ask questions in plain language. It can write SQL and Python code to analyze the data and produce charts, dashboards, HTML reports, and written summaries. It also works as a platform for building custom data agents and workflows.
- Analyze a CSV or Excel file and generate visual reports
- Connect to a database and produce a profiling report
- Ask business questions in natural language and have SQL written automatically
- Create and reuse SQL analysis skills for repeat tasks
Great for
- Analysts who want to ask questions of CSV or Excel files in plain language
- Teams exploring a database without writing SQL by hand
- Developers building AI-powered data assistants or agent workflows
- People who want to reuse domain-specific analysis steps as skills
Not ideal for
- Anyone who needs guaranteed-correct numbers without reviewing the generated SQL and code, since AI-written queries can be wrong
- Users without an LLM API key or local model setup, since the assistant needs a model to run
- Teams needing a small, simple tool, since the README describes a broad platform with many components
How to get it
- 1
Install with a package manager
You'll need: Python 3.10+ · uv (recommended) or pip
Install from PyPI with uv (recommended)
uv pip install dbgpt-appInstall from PyPI with pip
pip install dbgpt-appStart DB-GPT (first run opens a setup wizard)
dbgpt start - 2
Run it
- Install the dbgpt-app package with uv or pip
- Run dbgpt start and follow the interactive setup wizard to choose an LLM provider and enter an API key
- Open http://localhost:5670 in your browser
- Connect a CSV, Excel file, or database and ask a question in natural language
- 3
Get the most out of it
- Connect multiple sources (files, databases, knowledge bases) in one workspace to combine them in one analysis
- Load reusable skills for repeatable business analysis workflows
- Use the generated HTML reports and dashboards to share results
- Review the documentation at docs.dbgpt.cn for advanced profiles and configuration
Downloads
Strengths
- Covers the full path from connecting data to generating reports in one tool
- Runs code and tools in sandboxed environments, which the README presents as safer execution
- Released under the permissive MIT license with a large community (about 20,000 GitHub stars)
Trade-offs
- Output depends on the underlying AI model, so results need human checking
- The README is long and covers many features, so the setup and scope may take time to understand
- Some install and usage details are only in the external documentation site, not the README
How it compares
| Project | Stars | Type | Level | Updated | Best for |
|---|---|---|---|---|---|
| DB-GPT (this) | 20.2k | App | intermediate | 4 days ago | Analysts who want to ask questions of CSV or Excel files in plain language |
| langchain | 149k | Framework | intermediate | 4 days ago | Python developers building LLM-powered applications |
| llama_index | 52.8k | Library | intermediate | 10 days ago | Python developers building retrieval-augmented generation (RAG) apps over private documents |
| vanna | 23.8k | Library | intermediate | 9 months ago | Developers adding a natural-language data question box to an internal app |
| graphrag | 36.5k | Library | intermediate | 4 days ago | Researchers exploring graph-based retrieval over narrative or private text |
| LightRAG | 40.5k | Library | intermediate | 7 days ago | Developers building question-answering over their own document collections |
Stats synced from GitHub. The assessment above was drafted with AI from the project's README on 2026-10-29 and may miss recent changes — check the README before relying on a command. MIT licence: free to use, modify and distribute, including commercially, as long as you keep the copyright notice.