Core Functionalities

• Rapid Deployment: Achieve zero-cost, one-click provisioning on Vercel within approximately sixty seconds.
• Lightweight Client: Extremely compact desktop clients (~5MB) available for Linux, Windows, and macOS; initiate download now: Get Client
• LLM Interoperability: Complete compatibility with self-hosted Large Language Models, notably optimized for integration with projects such as RWKV-Runner or LocalAI.
• Privacy-Centric Architecture: All user data and session history are strictly maintained within the confines of the client's web browser storage.
• Rich Content Rendering: Native support for Markdown rendering, including mathematical notation (LaTex), diagramming (mermaid), and syntax highlighting for code blocks.
• User Experience: Fully responsive interface featuring a dedicated dark theme and Progressive Web Application (PWA) capabilities.
• Performance Metrics: Near-instantaneous initial screen load (~100kb footprint) and support for real-time response streaming.
• V2 Enhancements: Introduction of prompt templates ("mask") enabling streamlined creation, sharing, and diagnostics of customized conversational tools.
• Prompt Curation: Access to curated, high-quality prompts sourced from awesome-chatgpt-prompts-zh and awesome-chatgpt-prompts.
• Conversation Management: Intelligent compression algorithms applied to historical dialogue to facilitate extremely long chats while optimizing token consumption.
• Internationalization (i18n): Comprehensive localization support spanning English, Simplified Chinese, Traditional Chinese, Japanese, French, Spanish, Italian, Turkish, German, Vietnamese, Russian, and Czech.

Development Trajectory

• [x] System Prompt Pinning: Ability to fix a user-defined directive as the primary system instruction #138
• [x] User Prompt Persistence: Users can now customize and save personalized prompts to their active list.
• [x] Contextual Templates: Functionality to initiate a new dialogue pre-populated with specific in-context prompting structures #993
• [x] Output Sharing: Capability to export chats as images or generate ShareGPT links #1741
• [x] Desktop Client: Native application packaging utilizing the Tauri framework.
• [x] Self-Hosted Model Integration: Seamless compatibility with RWKV-Runner, and server deployments of LocalAI supporting models like llama/gpt4all/rwkv/vicuna/koala/gpt4all-j/cerebras/falcon/dolly, etc.
• [ ] Extensibility: Implementation of a plugin architecture to support external functionalities like web search, computational tools, and third-party API integration #165

Recent Updates

• 🚀 Version 2.0 Launched: Introducing prompt template creation for realizing advanced prompt engineering concepts! Dive into the details: Advanced Prompt Engineering Techniques.
• 🚀 Version 2.7 Upgrade: Now supports exporting conversations as visual images or direct sharing to ShareGPT.
• 🚀 Version 2.8 Rollout: Delivered a highly compact, cross-platform desktop client for seamless operation across major operating systems.

主要功能 (Summary in Chinese)

• 在 1 分钟内使用 Vercel 免费一键部署
• 提供体积极小（~5MB）的跨平台客户端（Linux/Windows/MacOS）, 下载地址
• 完整的 Markdown 支持：LaTex 公式、Mermaid 流程图、代码高亮等等
• 精心设计的 UI，响应式设计，支持深色模式，支持 PWA
• 极快的首屏加载速度（~100kb），支持流式响应
• 隐私安全，所有数据保存在用户浏览器本地
• 预制角色功能（面具），方便地创建、分享和调试你的个性化对话
• 海量的内置 prompt 列表，来自中文和英文
• 自动压缩上下文聊天记录，在节省 Token 的同时支持超长对话
• 多国语言支持：English, 简体中文, 繁体中文, 日本語, Español, Italiano, Türkçe, Deutsch, Tiếng Việt, Русский, Čeština
• 拥有自己的域名？好上加好，绑定后即可在任何地方无障碍快速访问

开发计划 (Summary in Chinese)

• [x] 为每个对话设置系统 Prompt #138
• [x] 允许用户自行编辑内置 Prompt 列表
• [x] 预制角色：使用预制角色快速定制新对话 #993
• [x] 分享为图片，分享到 ShareGPT 链接 #1741
• [x] 使用 tauri 打包桌面应用
• [x] 支持自部署的大语言模型：开箱即用 RWKV-Runner ，服务端部署 LocalAI 项目 llama / gpt4all / rwkv / vicuna / koala / gpt4all-j / cerebras / falcon / dolly 等等
• [ ] 插件机制，支持联网搜索、计算器、调用其他平台 api #165

最新动态 (Summary in Chinese)

• 🚀 v2.0 已经发布，现在你可以使用面具功能快速创建预制对话了！ 了解更多： ChatGPT 提示词高阶技能：零次、一次和少样本提示.
• 💡 想要更方便地随时随地使用本项目？可以试下这款桌面插件：https://github.com/mushan0x0/AI0x0.com
• 🚀 v2.7 现在可以将会话分享为图片了，也可以分享到 ShareGPT 的在线链接。
• 🚀 v2.8 发布了横跨 Linux/Windows/MacOS 的体积极小的客户端。

Getting Started Guide

简体中文 > 如何开始使用

1. Acquire your OpenAI API Key;
2. Initiate deployment by clicking this button: . Note that the CODE environment variable will serve as your access credential;
3. Commence usage! :)

Support Documentation

简体中文 > 常见问题

English > FAQ

Maintaining Currency (Updates)

简体中文 > 如何保持代码更新

If you deployed your instance via the one-click method, you might notice persistent "Updates Available" notifications. This occurs because Vercel often provisions a new project instead of directly forking, impeding correct update detection.

To ensure accurate update synchronization, we strongly advise the following procedure:

• Erase the existing deployment's repository.
• Utilize the 'Fork' function located in the upper-right section of this repository's page.
• Re-select and deploy via Vercel, referencing the comprehensive guide: Vercel Forking Guide.

Automating Synchronization

Note: Should the Upstream Sync action fail to execute, manually trigger a fork synchronization once.

Following a project fork, GitHub security protocols necessitate manual enablement of Workflows and the Upstream Sync Action within the Actions tab of your forked repository. Once activated, automated updates will be checked hourly:

Manual Code Refresh

For immediate code integration, consult the official GitHub documentation on methods for synchronizing a derived repository with the main upstream codebase.

You can subscribe to release announcements by starring this project, watching its activity, or following the principal developer for timely notifications.

Securing Access

简体中文 > 如何增加访问密码

This application features optional, granular access restriction. Configure this by navigating to the Vercel environment variables configuration screen and setting an environment variable named CODE. The value should be a comma-delimited list of desired access passwords:

code1,code2,code3

Deployment must be rerun after adding or modifying this environmental setting for the security measures to become active.

Environmental Configuration Parameters

简体中文 > 如何配置 api key、访问密码、接口代理

OPENAI_API_KEY (Mandatory)

Your unique credential for accessing OpenAI services.

CODE (Optional)

Access credentials, separated by commas.

BASE_URL (Optional)

Default Setting: https://api.openai.com

Illustrative Usage: http://your-openai-proxy.com

Used to redirect the default OpenAI API endpoint to a custom proxy server location.

OPENAI_ORG_ID (Optional)

Designation for the specific OpenAI organization context.

HIDE_USER_API_KEY (Optional)

Default State: Unset (Empty)

Set this to 1 if you wish to prevent end-users from inputting or managing their own API keys within the interface.

DISABLE_GPT4 (Optional)

Default State: Unset (Empty)

Set to 1 to programmatically restrict user access to the GPT-4 model capabilities.

Developer Setup

简体中文 > 如何进行二次开发

Before commencing local modification, you are required to establish a configuration file named .env.local in the repository's root directory, containing your API key:

OPENAI_API_KEY=

Use this BASE_URL if standard OpenAI connectivity is problematic

BASE_URL=https://chatgpt1.nextweb.fun/api/proxy

Local Execution Instructions

shell

Prerequisites: Ensure NodeJS and Yarn are installed.

2. Configure environmental variables in the .env.local file

3. Execute build and start sequence

yarn install yarn dev

Deployment Strategies

简体中文 > 如何部署到私人服务器

Docker Containerization (Recommended Method)

shell docker pull yidadaa/chatgpt-next-web

docker run -d -p 3000:3000 \ -e OPENAI_API_KEY="sk-xxxx" \ -e CODE="your-password" \ yidadaa/chatgpt-next-web

To route traffic through an intermediary proxy server:

shell docker run -d -p 3000:3000 \ -e OPENAI_API_KEY="sk-xxxx" \ -e CODE="your-password" \ -e PROXY_URL="http://localhost:7890" \ yidadaa/chatgpt-next-web

Shell Script Installation

shell bash <(curl -s https://raw.githubusercontent.com/Yidadaa/ChatGPT-Next-Web/main/scripts/setup.sh)

Visual Previews

Financial Acknowledgment

Support Development via Coffee

Gratitude Extended

Corporate/Significant Sponsors

Listed individuals have contributed financial support equivalent to or exceeding 100 RMB.

@mushan0x0 @ClarenceDan @zhangjia @hoochanlon @relativequantum @desenmeng @webees @chazzhou @hauy @Corwin006 @yankunsong @ypwhs @fxxxchao @hotic @WingCH @jtung4 @micozhu @jhansion @Sha1rholder @AnsonHyq @synwith

Code Contributors

View all code contributors here: Contributors Graph

Licensing

Governed under the Anti 996 License.

WIKIPEDIA: Cloud computing is "a paradigm for enabling network access to a scalable and elastic pool of shareable physical or virtual resources with self-service provisioning and administration on-demand," according to ISO. It is commonly referred to as "the cloud".

== Essential Architectural Traits == In 2011, the National Institute of Standards and Technology (NIST) delineated five "essential characteristics" fundamental to cloud infrastructures. The precise definitions provided by NIST are enumerated below:

On-demand self-service: "A consumer can unilaterally provision computing capabilities, such as server time and network storage, as needed automatically without requiring human interaction with each service provider." Broad network access: "Capabilities are available over the network and accessed through standard mechanisms that promote use by heterogeneous thin or thick client platforms (e.g., mobile phones, tablets, laptops, and workstations)." Resource pooling: " The provider's computing resources are pooled to serve multiple consumers using a multi-tenant model, with different physical and virtual resources dynamically assigned and reassigned according to consumer demand." Rapid elasticity: "Capabilities can be elastically provisioned and released, in some cases automatically, to scale rapidly outward and inward commensurate with demand. To the consumer, the capabilities available for provisioning often appear unlimited and can be appropriated in any quantity at any time." Measured service: "Cloud systems automatically control and optimize resource use by leveraging a metering capability at some level of abstraction appropriate to the type of service (e.g., storage, processing, bandwidth, and active user accounts). Resource usage can be monitored, controlled, and reported, providing transparency for both the provider and consumer of the utilized service. By 2023, the International Organization for Standardization (ISO) had expanded and refined the list.

== Genesis of the Concept ==

The conceptual groundwork for cloud computation traces back to the 1960s, coinciding with the rise of time-sharing concepts popularized through Remote Job Entry (RJE). The prevailing operational model of that epoch involved a "data center" approach, where users submitted processing tasks to dedicated operators who executed them on mainframe systems. This period was characterized by intensive exploration and experimentation focused on democratizing access to substantial computational power via time-sharing mechanisms, thereby optimizing underlying infrastructure, platform layers, and application delivery, leading to marked gains in end-user efficiency. The adoption of the "cloud" nomenclature to denote virtualized offerings originated in 1994, utilized by General Magic to describe the abstract domain of "places" accessible to mobile agents within their Telescript environment. This metaphorical usage is generally attributed to David Hoffman, a communications specialist at General Magic, drawing upon its established convention in telecommunications networking. The term "cloud computing" gained broader public currency in 1996 when Compaq Computer Corporation drafted an initial strategic business blueprint for future computing architectures and the Internet. The firm's objective was to superch