bentwng/chatgpt-web repository overview

⁠ChatGPT Web

中文⁠ | English⁠

⁠Introduction

Some unique features have been added:

[✓] Register & Login & Reset Password & 2FA

[✓] Sync chat history

[✓] Front-end page setting apikey

[✓] Custom Sensitive Words

[✓] Set unique prompts for each chat room

[✓] Users manager

[✓] Random Key

[✓] Conversation round limit & setting different limits by user & giftcards

[✓] Implement SSO login through the auth proxy feature (need to integrate a third-party authentication reverse proxy, it can support login protocols such as LDAP/OIDC/SAML)

⁠Screenshots

• ChatGPT Web⁠
  • Introduction⁠
  • Roadmap⁠
  • Prerequisites⁠
    • Node⁠
    • PNPM⁠
    • Fill in the Keys⁠
  • Install Dependencies⁠
    • Backend⁠
    • Frontend⁠
  • Run in Test Environment⁠
    • Backend Service⁠
    • Frontend Webpage⁠
  • Packaging⁠
    • Using Docker⁠
      • Docker Parameter Example⁠
      • Docker Build & Run⁠
      • Docker Compose⁠
    • Deployment with Railway⁠
      • Railway Environment Variables⁠
    • Manual packaging⁠
      • Backend service⁠
      • Frontend webpage⁠
  • Frequently Asked Questions⁠
  • Contributing⁠
  • Sponsorship⁠
  • License⁠

⁠Introduction

Supports dual models, provides two unofficial ChatGPT API methods:

Comparison:

1. ChatGPTAPI uses gpt-3.5-turbo-0301 to simulate ChatGPT through the official OpenAI completion API (the most reliable method, but it is not free and does not use models specifically tuned for chat).
2. ChatGPTUnofficialProxyAPI accesses ChatGPT's backend API via an unofficial proxy server to bypass Cloudflare (uses the real ChatGPT, is very lightweight, but depends on third-party servers and has rate limits).

Details⁠

Switching Methods:

1. Go to the service/.env.example file and copy the contents to the service/.env file.
2. For OpenAI API Key, fill in the OPENAI_API_KEY field (Get apiKey)⁠.
3. For Web API, fill in the OPENAI_ACCESS_TOKEN field (Get accessToken)⁠.
4. When both are present, OpenAI API Key takes precedence.

Reverse Proxy:

Available when using ChatGPTUnofficialProxyAPI.Details⁠

# service/.env
API_REVERSE_PROXY=

Environment Variables:

For all parameter variables, check here⁠ or see:

/service/.env

⁠Roadmap

[✓] Dual models

[✓] Multiple session storage and context logic

[✓] Formatting and beautifying code-like message types

[✓] Login or Register

[✓] Set API key and other information on the front-end page.

[✓] Data import and export

[✓] Save message to local image

[✓] Multilingual interface

[✓] Interface themes

[✗] More...

⁠Prerequisites

⁠Node

node requires version ^16 || ^18 || ^20 || ^22, and multiple local node versions can be managed using nvm⁠.

node -v

⁠PNPM

If you have not installed pnpm before:

npm install pnpm -g

⁠Fill in the Keys

Get Openai Api Key or accessToken and fill in the local environment variables jump⁠

# service/.env file

# OpenAI API Key - https://platform.openai.com/overview
OPENAI_API_KEY=

# change this to an `accessToken` extracted from the ChatGPT site's `https://chat.openai.com/api/auth/session` response
OPENAI_ACCESS_TOKEN=

⁠Install Dependencies

⁠Backend

Enter the /service folder and run the following command

pnpm install

⁠Frontend

Run the following command in the root directory

pnpm bootstrap

⁠Run in Test Environment

⁠Backend Service

Enter the /service folder and run the following command

pnpm start

⁠Frontend Webpage

Run the following command in the root directory

pnpm dev

⁠Packaging

⁠Using Docker

⁠Docker Parameter Example

• OPENAI_API_KEY one of two
• OPENAI_ACCESS_TOKEN one of two, OPENAI_API_KEY takes precedence when both are present
• OPENAI_API_BASE_URL optional, available when OPENAI_API_KEY is set
• OPENAI_API_MODEL ChatGPTAPI OR ChatGPTUnofficialProxyAPI
• API_REVERSE_PROXY optional, available when OPENAI_ACCESS_TOKEN is set Reference⁠
• AUTH_SECRET_KEY Access Password，optional
• TIMEOUT_MS timeout, in milliseconds, optional
• SOCKS_PROXY_HOST optional, effective with SOCKS_PROXY_PORT
• SOCKS_PROXY_PORT optional, effective with SOCKS_PROXY_HOST
• SOCKS_PROXY_USERNAME optional, effective with SOCKS_PROXY_HOST and SOCKS_PROXY_PORT
• SOCKS_PROXY_PASSWORD optional, effective with SOCKS_PROXY_HOST and SOCKS_PROXY_PORT
• HTTPS_PROXY optional, support http，https, socks5

⁠Docker Build & Run

GIT_COMMIT_HASH=`git rev-parse HEAD`
RELEASE_VERSION=`git branch --show-current`
docker build --build-arg GIT_COMMIT_HASH=${GIT_COMMIT_HASH} --build-arg RELEASE_VERSION=${RELEASE_VERSION} -t chatgpt-web .

# foreground operation
# If run mongodb in host machine, please use MONGODB_URL=mongodb://host.docker.internal:27017/chatgpt
docker run --name chatgpt-web --rm -it -p 127.0.0.1:3002:3002 --env OPENAI_API_KEY=your_api_key --env MONGODB_URL=your_mongodb_url chatgpt-web

# background operation
docker run --name chatgpt-web -d -p 127.0.0.1:3002:3002 --env OPENAI_API_KEY=your_api_key --env MONGODB_URL=your_mongodb_url chatgpt-web

# running address
http://localhost:3002/

⁠Docker Compose

Hub Address⁠

version: '3'

services:
  app:
    image: chatgptweb/chatgpt-web # always use latest, pull the tag image again when updating
    container_name: chatgptweb
    restart: unless-stopped
    ports:
      - 3002:3002
    depends_on:
      - database
    environment:
      TZ: Asia/Shanghai
      # Title for site
      SITE_TITLE: ChatGpt Web
      # access salt，optional Allow login if not empty.
      AUTH_SECRET_KEY: xxx
      # mongodb's connection string
      MONGODB_URL: 'mongodb://chatgpt:xxxx@database:27017'
      # After register enabled, Salt for password encryption
      PASSWORD_MD5_SALT: xxx
      # After register enabled, super administrator
      ROOT_USER: [email protected]
      # Allow anyone register, Must be turned on, otherwise administrators cannot register, can be turned off later.
      REGISTER_ENABLED: true
      # More configurations, register an administrator after running and set it in the administrator page.
    links:
      - database

  database:
    image: mongo
    container_name: chatgptweb-database
    restart: unless-stopped
    ports:
      - '27017:27017'
    expose:
      - '27017'
    volumes:
      - mongodb:/data/db
    environment:
      MONGO_INITDB_ROOT_USERNAME: chatgpt
      MONGO_INITDB_ROOT_PASSWORD: xxxx
      MONGO_INITDB_DATABASE: chatgpt

volumes:
  mongodb: {}

The OPENAI_API_BASE_URL is optional and only used when setting the OPENAI_API_KEY.

⁠Deployment with Railway

⁠Manual packaging

⁠Backend service

Copy the service folder to a server that has a node service environment.

# Install
pnpm install

# Build
pnpm build

# Run
pnpm prod

PS: You can also run pnpm start directly on the server without packaging.

⁠Frontend webpage

1. Refer to the root directory .env.example file content to create .env file, modify VITE_GLOB_API_URL in .env at the root directory to your actual backend interface address.
2. Run the following command in the root directory and then copy the files in the dist folder to the root directory of your website service.

Reference information⁠

pnpm build

⁠Frequently Asked Questions

Q: Why does Git always report an error when committing?

A: Because there is submission information verification, please follow the Commit Guidelines⁠.

Q: Where to change the request interface if only the frontend page is used?

A: The VITE_GLOB_API_URL field in the .env file at the root directory.

Q: All red when saving the file?

A: For vscode, please install the recommended plug-in of the project or manually install the Eslint plug-in.

Q: Why doesn't the frontend have a typewriter effect?

A: One possible reason is that after Nginx reverse proxying, buffering is turned on, and Nginx will try to buffer a certain amount of data from the backend before sending it to the browser. Please try adding proxy_buffering off; after the reverse proxy parameter and then reloading Nginx. Other web server configurations are similar.

Q: The content returned is incomplete?

A: There is a length limit for the content returned by the API each time. You can modify the VITE_GLOB_OPEN_LONG_REPLY field in the .env file under the root directory, set it to true, and rebuild the front-end to enable the long reply feature, which can return the full content. It should be noted that using this feature may bring more API usage fees.

⁠Auth Proxy Mode

Set env AUTH_PROXY_ENABLED=true can enable auth proxy mode.

After activating this feature, it is necessary to ensure that chatgpt-web can only be accessed through a reverse proxy.

Authentication is carried out by the reverse proxy, which then forwards the request with the header to identify the user identity. Default header name is X-Email, can custom config use set env AUTH_PROXY_HEADER_NAME.

Recommended for current IdP to use LDAP protocol, using authelia⁠

Recommended for current IdP to use OIDC protocol, using oauth2-proxy⁠

⁠Contributing

Please read the Contributing Guidelines⁠ before contributing.

Thanks to all the contributors!

⁠

⁠Star History

⁠Sponsorship

If you find this project helpful, please give me a star.

Thanks to DigitalOcean⁠ for sponsoring providing open-source credits used to run our infrastructure servers.

⁠

⁠License

MIT © github.com/chatgpt-web-dev Contributors⁠