akshare

资源分享：对于想了解更多财经数据与量化投研的小伙伴，推荐一个专注于财经数据和量化研究的知识社区。 该社区提供相关文档和视频学习资源，汇集了各类财经数据源和量化投研工具的使用经验。 有兴趣深入学习的朋友可点此了解更多，也推荐大家关注微信公众号【数据科学实战】。

重磅推荐：AKQuant 是一款专为 量化投研 (Quantitative Research) 打造的高性能量化回测框架。它以 Rust 铸造极速撮合内核， 以 Python 链接数据与 AI 生态，旨在为量化投资者提供可靠高效的量化投研解决方案。参见AKQuant

工具推荐：期魔方是一款本地化期货量化分析工具，适合数据分析爱好者使用。无需复杂部署，支持数据分析和机器学习功能，研究功能免费开放。 如需了解更多信息可访问期魔方。

AKShare requires Python(64 bit) 3.11 or higher and aims to simplify the process of fetching financial data.

Write less, get more!

• Documentation: 中文文档
• MCP server for AI assistants: AKTools lets Claude, Cursor and other MCP clients fetch AKShare data in natural language — see AI Assistants (MCP)
• HTTP API: AKTools also serves AKShare over HTTP for other programming languages

AKShare is a Python library and does not run an MCP server itself. Its companion project AKTools ships one — registered in the official MCP Registry as io.github.akfamily/aktools — so AI assistants such as Claude Code, Claude Desktop and Cursor can look up and call every AKShare interface from a natural-language question, e.g. "贵州茅台 2024 年 10 月 8 日到 11 日的前复权日线行情".

No separate install is needed beyond uv:

claude mcp add aktools -- uvx --from aktools aktools-mcp

Claude Desktop, Cursor and other clients:

{
  "mcpServers": {
    "aktools": {
      "command": "uvx",
      "args": ["--from", "aktools", "aktools-mcp"]
    }
  }
}

Instead of registering 1000+ tools, it exposes three generic ones that the assistant chains together:

A running python -m aktools server also exposes the same tools over Streamable HTTP at http://127.0.0.1:8080/mcp. See the MCP guide (中文) for HTTP mode, proxies, access tokens and LAN access.

uv pip install akshare --upgrade

See the uv environment guide for installing uv and creating a virtual environment.

pip install akshare --upgrade

uv pip install akshare --upgrade --default-index https://pypi.tuna.tsinghua.edu.cn/simple
# or
pip install akshare --upgrade -i https://pypi.tuna.tsinghua.edu.cn/simple

Code:

import akshare as ak

stock_zh_a_hist_df = ak.stock_zh_a_hist(symbol="000001", period="daily", start_date="20170301", end_date="20231022", adjust="")
print(stock_zh_a_hist_df)

Output:

      日期          开盘   收盘    最高  ...  振幅   涨跌幅  涨跌额  换手率
0     2017-03-01   9.49   9.49   9.55  ...  0.84  0.11  0.01  0.21
1     2017-03-02   9.51   9.43   9.54  ...  1.26 -0.63 -0.06  0.24
2     2017-03-03   9.41   9.40   9.43  ...  0.74 -0.32 -0.03  0.20
3     2017-03-06   9.40   9.45   9.46  ...  0.74  0.53  0.05  0.24
4     2017-03-07   9.44   9.45   9.46  ...  0.63  0.00  0.00  0.17
          ...    ...    ...    ...  ...   ...   ...   ...   ...
1610  2023-10-16  11.00  11.01  11.03  ...  0.73  0.09  0.01  0.26
1611  2023-10-17  11.01  11.02  11.05  ...  0.82  0.09  0.01  0.25
1612  2023-10-18  10.99  10.95  11.02  ...  1.00 -0.64 -0.07  0.34
1613  2023-10-19  10.91  10.60  10.92  ...  3.01 -3.20 -0.35  0.61
1614  2023-10-20  10.55  10.60  10.67  ...  1.51  0.00  0.00  0.27
[1615 rows x 11 columns]

AKShare ships an offline interface registry, so you can look up an interface by keyword without leaving Python and without any network request. This is especially handy for LLM-driven programs that need to resolve a description into a callable interface name.

Code:

import akshare as ak

search_df = ak.search("可转债 实时行情", limit=5)
print(search_df)

# full metadata of a single interface: params, output columns and a usage example
interface_info_dict = ak.interface_info("bond_cb_jsl")
print(interface_info_dict)

Output:

                      接口名    类目                                                       描述  有无文档   匹配分
0             bond_cb_jsl  bond  集思录可转债实时数据，包含行情数据（涨跌幅，成交量和换手率等）及可转债基本信息（转股价，溢价率和到期收益率等）  True  27.0
1      bond_zh_hs_cov_min  bond                                           东方财富网-可转债-分时行情  True  20.0
2  bond_zh_hs_cov_pre_min  bond                                      东方财富网-可转债-分时行情-盘前分时  True  20.0
3     bond_cov_comparison  bond                                   东方财富网-行情中心-债券市场-可转债比价表  True  17.0
4        fund_etf_spot_em  fund                                            东方财富-ETF 实时行情  True  17.0

Passing a full interface name always ranks that interface first, which makes it a reliable way to confirm a half-remembered name. Note this is keyword matching rather than semantic search — see the interface search docs for what it can and cannot do.

Code:

import akshare as ak
import mplfinance as mpf  # Please install mplfinance as follows: pip install mplfinance

stock_us_daily_df = ak.stock_us_daily(symbol="AAPL", adjust="qfq")
stock_us_daily_df = stock_us_daily_df.set_index(["date"])
stock_us_daily_df = stock_us_daily_df["2020-04-01": "2020-04-29"]
mpf.plot(stock_us_daily_df, type="candle", mav=(3, 6, 9), volume=True, show_nontrading=False)

• Ease of use: Just one line code to fetch the data;
• Extensible: Easy to customize your own code with other application;
• Powerful: Python ecosystem.

1. Overview
2. Installation
3. Interface Search
4. AI Assistants (MCP)
5. Tutorial
6. Data Dict
7. Indicator
8. FAQ
9. Trade

AKShare is still under development, feel free to open issues and pull requests:

• Report or fix bugs
• Require or publish interface
• Write or fix documentation
• Add test cases

See the contributing guide before opening a pull request.

Notice: We use Ruff to format the code

1. All data provided by AKShare is just for academic research purpose;
2. The data provided by AKShare is for reference only and does not constitute any investment proposal;
3. Any investor based on AKShare research should pay more attention to data risk;
4. AKShare will insist on providing open-source financial data;
5. Based on some uncontrollable factors, some data interfaces in AKShare may be removed;
6. Please follow the relevant open-source protocol used by AKShare;
7. Provide HTTP API and MCP server for other programming languages and AI assistants: AKTools.

Use the badge in your project's README.md:

[![Data: akshare](https://img.shields.io/badge/Data%20Science-AKShare-green)](https://github.com/akfamily/akshare)

Using the badge in README.rst:

.. image:: https://img.shields.io/badge/Data%20Science-AKShare-green
    :target: https://github.com/akfamily/akshare

Looks like this:

Please use this bibtex if you want to cite this repository in your publications:

@misc{akshare,
    author = {Albert King and Yaojie Zhang},
    title = {AKShare},
    year = {2022},
    publisher = {GitHub},
    journal = {GitHub repository},
    howpublished = {\url{https://github.com/akfamily/akshare}},
}

Special thanks FuShare for the opportunity of learning from the project;

Special thanks TuShare for the opportunity of learning from the project;

Thanks for the data provided by 东方财富网站;

Thanks for the data provided by 新浪财经网站;

Thanks for the data provided by 金十数据网站;

Thanks for the data provided by 生意社网站;

Thanks for the data provided by 中国银行间市场交易商协会网站;

Thanks for the data provided by 99期货网站;

Thanks for the data provided by 中国外汇交易中心暨全国银行间同业拆借中心网站;

Thanks for the data provided by 和讯财经网站;

Thanks for the data provided by DACHENG-XIU 网站;

Thanks for the data provided by 上海证券交易所网站;

Thanks for the data provided by 深圳证券交易所网站;

Thanks for the data provided by 北京证券交易所网站;

Thanks for the data provided by 中国金融期货交易所网站;

Thanks for the data provided by 上海期货交易所网站;

Thanks for the data provided by 大连商品交易所网站;

Thanks for the data provided by 郑州商品交易所网站;

Thanks for the data provided by 上海国际能源交易中心网站;

Thanks for the data provided by Timeanddate 网站;

Thanks for the data provided by 河北省空气质量预报信息发布系统网站;

Thanks for the data provided by Economic Policy Uncertainty 网站;

Thanks for the data provided by 申万指数网站；

Thanks for the data provided by 真气网网站;

Thanks for the data provided by 财富网站;

Thanks for the data provided by 中国证券投资基金业协会网站;

Thanks for the data provided by Expatistan 网站;

Thanks for the data provided by 北京市碳排放权电子交易平台网站;

Thanks for the data provided by 国家金融与发展实验室网站;

Thanks for the data provided by 义乌小商品指数网站;

Thanks for the data provided by 百度迁徙网站;

Thanks for the data provided by 思知网站;

Thanks for the data provided by Currencyscoop 网站;

Thanks for the data provided by 新加坡交易所网站;