Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/dataease/SQLBot/llms.txt

Use this file to discover all available pages before exploring further.

SQLBot is an intelligent data query system built by the DataEase open-source team. It connects large language models with your databases through a RAG pipeline, letting users ask questions in plain language and receive SQL-backed results as tables or interactive charts — no SQL knowledge required. Designed to improve over time, SQLBot gets more accurate the more it is used.

​
Key features

Text-to-SQL via LLMs + RAG

Natural language questions are converted to SQL using a retrieval-augmented pipeline that feeds the LLM only the schema context it needs, improving accuracy and reducing hallucinations.

ChatBI

A conversational interface lets users ask follow-up questions, regenerate answers, request analysis, and get recommended next questions — all in a single chat session.

Multi-tenant workspaces

Workspace-level resource isolation builds clear data boundaries. Fine-grained permission controls ensure every user only sees the data they are allowed to access.

12+ LLM providers

Works out of the box with OpenAI, DeepSeek, Gemini, Kimi, Qianfan, Tencent Hunyuan, iFlytek Spark, and any OpenAI-compatible endpoint.

Multiple database connectors

Connects to MySQL, PostgreSQL, SQL Server, Oracle, ClickHouse, Elasticsearch, Redshift, Hive, DM, and Excel/CSV files.

Web embedding and MCP

Embed SQLBot as a chat widget with web or pop-up embedding. Expose it as an MCP server so AI agents in n8n, Dify, MaxKB, or DataEase can query your databases directly.

​
Supported LLM providers

SQLBot supports 12 LLM providers, all communicating over the OpenAI-compatible API format.
ProviderAPI compatibility
Alibaba Cloud BailianOpenAI compatible
QianfanOpenAI compatible
DeepSeekOpenAI compatible
Tencent HunyuanOpenAI compatible
iFlytek SparkOpenAI compatible
GeminiOpenAI compatible
OpenAINative
KimiOpenAI compatible
Tencent CloudOpenAI compatible
Volcano EngineOpenAI compatible
MiniMaxOpenAI compatible
Generic OpenAI compatibleCustom base URL
The generic OpenAI-compatible provider lets you connect any locally hosted or third-party model that exposes an OpenAI-style API, including Ollama, LM Studio, and vLLM.

​
Accuracy over time

SQLBot provides three mechanisms that improve query accuracy as you use it:
  • Terminology library — Define domain-specific vocabulary and synonyms (e.g., “revenue” maps to the net_sales column). SQLBot retrieves relevant terms via embedding similarity and injects them into the prompt.
  • Data training — Store question-to-SQL example pairs. These are embedded and retrieved at query time so the LLM can use proven patterns from your business context.
  • Table relations — Declare how tables join together. SQLBot includes this metadata when constructing prompts, reducing incorrect cross-joins and missing foreign keys.

​
Ready to get started?

Quickstart

Deploy SQLBot with Docker and run your first query in minutes.

How it works

Understand the RAG + LLM pipeline that powers SQLBot.

Build docs developers (and LLMs) love

Get started for freeTalk to us