ChatGPT Web

English / 简体中文

A commercially-viable ChatGpt web application built with React.

可部署商业化的 ChatGpt 网页应用。

Proxy Demo / Business Demo / Issues / Buy Me a Coffee

🤖 Major Function

• The user system can impose relevant restrictions on usage
• Carefully designed UI, responsive design.
• Extremely fast first screen loading speed（~100kb）
• Massive built-in prompt list fromChinese AND English
• One click export of chat records, complete Markdown support.
• Support for custom API addresses（example：openAI / API2D）

🎮 Start Using

Node

Node requires version ^ 16 | | ^ 18 | | ^ 19 (node>=16), and NVM can be used to manage multiple local node versions.

# View node version
node -v

# View npm version
npm -v

# View yarn version
yarn -v

1.First Fork this project, then clone it locally.

git clone https://github.com/zihjiang/ChatGpt-Web.git

2.Installation dependencies

yarn install

3.Run

yarn dev

4.Build

yarn build

⛺️ Environment Variable

Most configuration items in this project are set through environment variables.

VITE_APP_REQUEST_HOST

Request the Host address of the server.

VITE_APP_TITLE

Chat Web title.

VITE_APP_LOGO

Chat Web Logo。

VITE_APP_MODE

Optional application mode: business mode proxy pattern mixed mode.

VITE_APP_AI_BASE_URL

Built in request address and KEY settings

VITE_APP_AI_KEYS

The key used by Api requests, supporting multiple keys separated by commas (,)

🚧 Develop

It is strongly not recommended to develop or deploy locally. Due to technical reasons, it is difficult to configure OpenAI API proxies locally, unless you can guarantee direct connection to the OpenAI server.

Local development

1. Install Nodejs and Yarn, please consult ChatGPT for specific details;
2. Execute yarn install && yarn dev.

Server side

1. Currently, the server is not yet fully developed, so it is currently not open source.
2. Front end request server's interface document You can develop according to this interface document.

🎯 Arrange

Simply upload the packaged dist directory to the server. The WEB project temporarily does not directly access the OpenAI API and does not require a server address.

🧘 Contributor

See project contributor list

📋 License