A Model Context Protocol (MCP) server that enables natural language queries to databases
powered by XiYan-SQL, SOTA of text-to-sql on open benchmarks
💻 XiYan-mcp-server |
🌐 XiYan-SQL |
📖 Arxiv |
📄 PapersWithCode
🤗 HuggingFace |
🤖 ModelScope |
🌕 析言GBI
English | 中文 | 日本語
Ding Group钉钉群|
Follow me on Weibo
- 🌐 Fetch data by natural language through XiYanSQL
- 🤖 Support general LLMs (GPT,qwenmax), Text-to-SQL SOTA model
- 💻 Support pure local mode (high security!)
- 📝 Support MySQL and PostgreSQL.
- 🖱️ List available tables as resources
- 🔧 Read table contents
There are two ways to integrate this server in your project, as shown below: The left is remote mode, which is the default mode. It requires an API key to access the xiyanSQL-qwencoder-32B model from service provider (see Configuration). Another mode is local mode, which is more secure. It does not require the API key.
The following figure illustrates the performance of the XiYan MCP server as measured by the MCPBench benchmark. The XiYan MCP server demonstrates superior performance compared to both the MySQL MCP server and the PostgreSQL MCP server, achieving a lead of 2-22 percentage points. The detailed experiment results can be found at MCPBench and the report "Evaluation Report on MCP Servers".
-
The tool
get_data
provides a natural language interface for retrieving data from a database. This server will convert the input natural language into SQL using a built-in model and call the database to return the query results. -
The
{dialect}://{table_name}
resource allows obtaining a portion of sample data from the database for model reference when a specific table_name is specified. -
The
{dialect}://
resource will list the names of the current databases
Python 3.11+ is required. You can install the server through pip, and it will install the latest version:
pip install xiyan-mcp-server
If you want to install the development version from source, you can install from source code on github:
pip install git+https://github.com/XGenerationLab/xiyan_mcp_server.git
See @XGenerationLab/xiyan_mcp_server
Not fully tested.
You need a YAML config file to configure the server. A default config file is provided in config_demo.yml which looks like this:
mcp:
transport: "stdio"
model:
name: "XGenerationLab/XiYanSQL-QwenCoder-32B-2412"
key: ""
url: "https://api-inference.modelscope.cn/v1/"
database:
host: "localhost"
port: 3306
user: "root"
password: ""
database: ""
You can set the transport protocol to stdio
or sse
.
For stdio protocol, you can set just like this:
mcp:
transport: "stdio"
For sse protocol, you can set mcp config as below:
mcp:
transport: "sse"
port: 8000
log_level: "INFO"
The default port is 8000
. You can change the port if needed.
The default log level is ERROR
. We recommend to set log level to INFO
for more detailed information.
Other configurations like debug
, host
, sse_path
, message_path
can be customized as well, but normally you don't need to modify them.
Name
is the name of the model to use, key
is the API key of the model, url
is the API url of the model. We support following models.
versions | general LLMs(GPT,qwenmax) | SOTA model by Modelscope | SOTA model by Dashscope | Local LLMs |
---|---|---|---|---|
description | basic, easy to use | best performance, stable, recommand | best performance, for trial | slow, high-security |
name | the official model name (e.g. gpt-3.5-turbo,qwen-max) | XGenerationLab/XiYanSQL-QwenCoder-32B-2412 | xiyansql-qwencoder-32b | xiyansql-qwencoder-3b |
key | the API key of the service provider (e.g. OpenAI, Alibaba Cloud) | the API key of modelscope | the API key via email | "" |
url | the endpoint of the service provider (e.g."https://api.openai.com/v1") | https://api-inference.modelscope.cn/v1/ | https://xiyan-stream.biz.aliyun.com/service/api/xiyan-sql | http://localhost:5090 |
If you want to use the general LLMs, e.g. gpt3.5, you can directly config like this:
model:
name: "gpt-3.5-turbo"
key: "YOUR KEY "
url: "https://api.openai.com/v1"
database:
If you want to use Qwen from Alibaba, e.g. Qwen-max, you can use following config:
model:
name: "qwen-max"
key: "YOUR KEY "
url: "https://dashscope.aliyuncs.com/compatible-mode/v1"
database:
We recommend the XiYanSQL-qwencoder-32B (https://github.com/XGenerationLab/XiYanSQL-QwenCoder), which is the SOTA model in text-to-sql, see Bird benchmark. There are two ways to use the model. You can use either of them. (1) Modelscope, (2) Alibaba Cloud DashScope.
You need to apply a key
of API-inference from Modelscope, https://www.modelscope.cn/docs/model-service/API-Inference/intro
Then you can use the following config:
model:
name: "XGenerationLab/XiYanSQL-QwenCoder-32B-2412"
key: ""
url: "https://api-inference.modelscope.cn/v1/"
Read our model description for more details.
We deployed the model on Alibaba Cloud DashScope, so you need to set the following environment variables:
Send me your email to get the key
. ( [email protected] )
In the email, please attach the following information:
name: "YOUR NAME",
email: "YOUR EMAIL",
organization: "your college or Company or Organization"
We will send you a key
according to your email. And you can fill the key
in the yml file.
The key
will be expired by 1 month or 200 queries or other legal restrictions.
model:
name: "xiyansql-qwencoder-32b"
key: "KEY"
url: "https://xiyan-stream.biz.aliyun.com/service/api/xiyan-sql"
Note: this model service is just for trial, if you need to use it in production, please contact us.
Alternatively, you can also deploy the model XiYanSQL-qwencoder-32B on your own server. See Local Model for more details.
host
, port
, user
, password
, database
are the connection information of the database.
You can use local or any remote databases. Now we support MySQL and PostgreSQL(more dialects soon).
database:
host: "localhost"
port: 3306
user: "root"
password: ""
database: ""
Step 1: Install Python packages
pip install psycopg2
Step 2: prepare the config.yml like this:
database:
dialect: "postgresql"
host: "localhost"
port: 5432
user: ""
password: ""
database: ""
Note that dialect
should be postgresql
for postgresql.
If you want to launch server with sse
, you have to run the following command in a terminal:
YML=path/to/yml python -m xiyan_mcp_server
Then you should see the information on http://localhost:8000/sse in your browser. (Defaultly, change if your mcp server runs on other host/port)
Otherwise, if you use stdio
transport protocol, you usually declare the mcp server command in specific mcp application instead of launching it in a terminal.
However, you can still debug with this command if needed.
Add this in your Claude Desktop config file, ref Claude Desktop config example
{
"mcpServers": {
"xiyan-mcp-server": {
"command": "/xxx/python",
"args": [
"-m",
"xiyan_mcp_server"
],
"env": {
"YML": "PATH/TO/YML"
}
}
}
}
Please note that the Python command here requires the complete path to the Python executable (/xxx/python
); otherwise, the Python interpreter cannot be found. You can determine this path by using the command which python
. The same applies to other applications as well.
Claude Desktop currently does not support the SSE transport protocol.
Prepare the config like Claude Desktop
If you use stdio
, add following command in the config, ref Goose config example
env YML=path/to/yml /xxx/python -m xiyan_mcp_server
Otherwise, if you use sse
, change Type to SSE
and set the endpoint to http://127.0.0.1:8000/sse
Use the similar command as follows.
For stdio
:
{
"mcpServers": {
"xiyan-mcp-server": {
"command": "/xxx/python",
"args": [
"-m",
"xiyan_mcp_server"
],
"env": {
"YML": "path/to/yml"
}
}
}
}
For sse
:
{
"mcpServers": {
"xiyan_mcp_server_1": {
"url": "http://localhost:8000/sse"
}
}
}
Add following in command:
/xxx/python -m xiyan_mcp_server
Add an env: key is YML and value is the path to your yml file. Ref Witsy config example
Contact us: Ding Group钉钉群| Follow me on Weibo
If you find our work helpful, feel free to give us a cite.
@article{xiyansql,
title={A Preview of XiYan-SQL: A Multi-Generator Ensemble Framework for Text-to-SQL},
author={Yingqi Gao and Yifu Liu and Xiaoxia Li and Xiaorong Shi and Yin Zhu and Yiming Wang and Shiqi Li and Wei Li and Yuntao Hong and Zhiling Luo and Jinyang Gao and Liyu Mou and Yu Li},
year={2024},
journal={arXiv preprint arXiv:2411.08599},
url={https://arxiv.org/abs/2411.08599},
primaryClass={cs.AI}
}