Top Related Projects
AI agent stdlib that works with any LLM and TypeScript AI SDK.
🔮 ChatGPT Desktop Application (Mac, Windows and Linux)
Use ChatGPT On Wechat via wechaty
用 Express 和 Vue3 搭建的 ChatGPT 演示网页
A cross-platform ChatGPT/Gemini UI (Web / PWA / Linux / Win / MacOS). 一键拥有你自己的跨平台 ChatGPT/Gemini 应用。
Quick Overview
The ourongxing/chatgpt-vercel repository is a project that deploys ChatGPT on Vercel, providing a web-based interface for interacting with OpenAI's language model. It offers a customizable and easily deployable solution for creating a ChatGPT-like experience on your own domain.
Pros
- Easy deployment on Vercel with minimal configuration
- Customizable UI and functionality to suit specific needs
- Supports multiple API endpoints, including Azure OpenAI
- Implements rate limiting and request throttling for better resource management
Cons
- Requires an OpenAI API key, which may involve costs
- Limited to the capabilities of the underlying OpenAI model
- May require additional configuration for advanced features
- Potential for high usage costs if not properly managed
Getting Started
- Fork the repository on GitHub
- Sign up for a Vercel account if you don't have one
- Create a new project on Vercel and connect it to your forked repository
- Set the following environment variables in your Vercel project settings:
OPENAI_API_KEY
: Your OpenAI API keyCODE
: (Optional) Set a password for accessing the chat interface
- Deploy the project on Vercel
- Access your deployed ChatGPT interface at the provided Vercel URL
For more detailed instructions and customization options, refer to the repository's README.md file.
Competitor Comparisons
AI agent stdlib that works with any LLM and TypeScript AI SDK.
Pros of agentic
- More comprehensive AI agent framework with support for multiple models and tools
- Focuses on building autonomous AI agents with goal-oriented behavior
- Provides a flexible architecture for creating complex AI workflows
Cons of agentic
- Steeper learning curve due to its more advanced features and concepts
- May be overkill for simple chatbot applications
- Requires more setup and configuration compared to chatgpt-vercel
Code comparison
chatgpt-vercel:
export default async function handler(req: NextRequest) {
const { prompt, key, temperature } = (await req.json()) as ChatRequest
const apiKey = key || process.env.OPENAI_API_KEY
// ... (API call and response handling)
}
agentic:
const agent = new Agent({
name: 'My Agent',
goal: 'Accomplish a specific task',
tools: [new WebBrowser(), new Calculator()],
model: new OpenAIChat({ apiKey: process.env.OPENAI_API_KEY })
})
const result = await agent.run()
The code snippets highlight the difference in complexity and focus between the two projects. chatgpt-vercel provides a simpler API handler for chat requests, while agentic offers a more sophisticated agent setup with various tools and goal-oriented behavior.
🔮 ChatGPT Desktop Application (Mac, Windows and Linux)
Pros of ChatGPT
- Desktop application with cross-platform support (Windows, macOS, Linux)
- More feature-rich, including prompt library, export options, and system tray support
- Regular updates and active development
Cons of ChatGPT
- Larger application size due to being a desktop app
- Requires installation and updates, unlike web-based solutions
- May have a steeper learning curve for new users due to additional features
Code Comparison
ChatGPT (Tauri-based desktop app):
#[tauri::command]
fn get_data() -> String {
// Desktop-specific implementation
}
chatgpt-vercel (Web-based):
export async function getData() {
// Web-based implementation
}
The code comparison highlights the fundamental difference in architecture between the two projects. ChatGPT uses Rust with Tauri for desktop app development, while chatgpt-vercel employs JavaScript for a web-based solution. This difference affects deployment, user experience, and potential feature sets.
ChatGPT offers a more comprehensive, native-like experience with offline capabilities and system integration. In contrast, chatgpt-vercel provides a lightweight, easily accessible web interface without installation requirements. The choice between the two depends on user preferences, target platforms, and desired features.
Use ChatGPT On Wechat via wechaty
Pros of wechat-chatgpt
- Integrates directly with WeChat, allowing for seamless ChatGPT interactions within the popular messaging platform
- Supports multiple ChatGPT accounts for load balancing and improved reliability
- Includes features like message retry and error handling for a more robust user experience
Cons of wechat-chatgpt
- More complex setup process due to WeChat integration requirements
- Limited to WeChat platform, reducing flexibility for users on other platforms
- May require ongoing maintenance to keep up with WeChat API changes
Code Comparison
wechat-chatgpt:
const { WechatyBuilder } = require('wechaty')
const qrcodeTerminal = require('qrcode-terminal')
const bot = WechatyBuilder.build()
bot.on('scan', (qrcode, status) => {
qrcodeTerminal.generate(qrcode, { small: true })
})
chatgpt-vercel:
import { ChatGPTAPI } from 'chatgpt'
import { createParser } from 'eventsource-parser'
export default async function handler(req, res) {
const api = new ChatGPTAPI({ apiKey: process.env.OPENAI_API_KEY })
const response = await api.sendMessage(req.body.message)
}
The code snippets highlight the different focus areas of each project. wechat-chatgpt emphasizes WeChat integration, while chatgpt-vercel provides a more straightforward API interaction with ChatGPT, making it easier to deploy on Vercel's serverless platform.
用 Express 和 Vue3 搭建的 ChatGPT 演示网页
Pros of chatgpt-web
- More comprehensive features, including conversation management and export functionality
- Better suited for self-hosting and customization
- Supports multiple API endpoints and custom API keys
Cons of chatgpt-web
- More complex setup and configuration process
- Requires additional server resources due to its expanded feature set
- May have a steeper learning curve for users unfamiliar with Vue.js and Node.js
Code Comparison
chatgpt-web (Vue.js component):
<template>
<div class="chat-message" :class="{ 'chat-message-user': inversion }">
<div class="chat-message-container">
<div class="chat-message-avatar">
<img :src="avatar" alt="avatar">
</div>
<div class="chat-message-content">
<template v-if="!inversion">
<div class="chat-message-text" v-html="text"></div>
</template>
<template v-else>
<div class="chat-message-text">{{ text }}</div>
</template>
</div>
</div>
</div>
</template>
chatgpt-vercel (React component):
export function ChatMessage(props: {
message: Message
showWordCount?: boolean
showTokenCount?: boolean
showModelName?: boolean
hightlightCodeBlock?: boolean
}) {
return (
<div className={`flex flex-col ${props.message.role === 'user' ? 'items-end' : 'items-start'}`}>
<div className={`flex items-center ${props.message.role === 'user' ? 'flex-row-reverse' : ''}`}>
<Avatar role={props.message.role} />
<div className={`px-3 py-2 rounded-lg ${props.message.role === 'user' ? 'bg-[#10a37f] text-white' : 'bg-[#f4f6f8] text-black'}`}>
{props.message.content}
</div>
</div>
</div>
)
}
A cross-platform ChatGPT/Gemini UI (Web / PWA / Linux / Win / MacOS). 一键拥有你自己的跨平台 ChatGPT/Gemini 应用。
Pros of ChatGPT-Next-Web
- More feature-rich with advanced functionalities like custom prompts and plugins
- Active development with frequent updates and improvements
- Larger community support and contributions
Cons of ChatGPT-Next-Web
- More complex setup and configuration process
- Higher resource requirements due to additional features
Code Comparison
ChatGPT-Next-Web:
export function Markdown(props: { content: string }) {
return (
<ReactMarkdown
remarkPlugins={[remarkGfm, remarkMath]}
rehypePlugins={[rehypeMathjax, rehypeHighlight]}
components={{
code({ node, inline, className, children, ...props }) {
// ... (code implementation)
},
}}
{...props}
>
{props.content}
</ReactMarkdown>
);
}
chatgpt-vercel:
export default function Markdown({ content }: { content: string }) {
return (
<ReactMarkdown
remarkPlugins={[remarkGfm]}
components={{
code({ node, inline, className, children, ...props }) {
// ... (code implementation)
},
}}
>
{content}
</ReactMarkdown>
);
}
The code comparison shows that ChatGPT-Next-Web includes additional plugins for enhanced markdown rendering, such as math support and syntax highlighting, while chatgpt-vercel uses a simpler implementation with fewer plugins.
Convert designs to code with AI
Introducing Visual Copilot: A new AI model to turn Figma designs to high quality code using your components.
Try Visual CopilotREADME
ChatGPT-Vercel
æ¬é¡¹ç®åºäº chatgpt-demo å¼åãæ£å¼çæ¬ï¼v1.0.0ï¼ä½¿ç¨ Solid-Start æ¡æ¶ï¼å¹¶ä¸ä¸åå¼å¯ SSRã请éæ°è®¾ç½®ç¯å¢åéã
å¨çº¿é¢è§:
- vercel.app 被å¢ã
ç¹ç¹
- UI ä¼é ç®æ´çåæ¶åè½å°½å¯è½å¼ºå¤§ï¼ç»è满满ï¼çä½ åç°ãåæ¶æ»¡è¶³å°ç½å大佬使ç¨ã
- æ¯æ PWAã
- æ¯æ Prompts é¢è®¾ï¼ä½¿ç¨ ç©ºæ ¼ æè / æç´¢ï¼å¯ä»¥æ¨¡ç³å¹é ï¼çªåºæ¾ç¤ºå¹é çæåã
- æ¯æå¤ä¸ªå¯¹è¯ï¼è天ï¼ï¼å¹¶ä¸æ¯ä¸ªå¯¹è¯é½å¯ä»¥åç¬è®¾ç½®ï¼èªç±è®¾ç½®è§è²ï¼ä¸å对è¯å¯ä»¥éè¿ URL ç´è¾¾ãä½¿ç¨ ç©ºæ ¼ç©ºæ ¼ æè // æç´¢ï¼å¯ä»¥æ¨¡ç³å¹é ï¼æç´¢åå²æ¶æ¯ã
- æ¯æ导åºåå¯¼å ¥å¯¹è¯å设置ã
- å®æ¶æ¥çå½å对è¯ä¼æ¶èç token 以ååçç tokenï¼åæ¶ææã
- æ¯æå¤ç§ä½é¢æ¥è¯¢æ¹å¼ï¼æ¯æå®æ¶æ¥è¯¢å 置账å·ä½é¢ï¼å¹¶ä¸åéå°å¾®ä¿¡ã
- æ¯æ URL queryï¼å½ä½æç´¢å¼æ使ç¨ï¼ææ对è¯é½æ¯æãæ¯å¦
url?q=keyword
æèurl/session/xxxxx?q=keyword
使ç¨æå·§
- ææä¸ä¸æï¼åé¤æ¥éçé¨å对è¯ãå¦æå ³éè¿ç»å¯¹è¯ï¼å°åªç»è®¡éå®äºçé¨å对è¯ãè¿ä¸ªæ¦å¿µå¾å ³é®ã
- éå®å¯¹è¯ï¼ç¹å»æ¯æ¡æ¶æ¯åé¢ç头åï¼å°±å¯ä»¥éå®äºãéå®æä»ä¹å¥½å¤ï¼æ¸
空对è¯æ¶ï¼éå®äºçä¸ä¼å é¤ãå
³éè¿ç»å¯¹è¯æ¶ï¼å§ç»ä¼åééå®äºç对è¯ï¼ç¸å½äºè§è²è®¾å®ãå
³é
è®°å½å¯¹è¯å 容
åï¼å§ç»ä¼è®°å½éå®äºç对è¯ï¼ä¸æ¬¡å·æ°ä»ç¶åå¨ã - Open AI Key è¦æä¹è·å¾ï¼æ³¨å OpenAI çå¸å·ï¼ç¶å çæ Key å°±è¡äºãç°å¨æ³¨åå°±é 5 ç¾å ï¼å¯ä»¥ç¨ä¸ä¸¤ä¸ªæãå«æ³¨å麻ç¦ï¼å¯ä»¥ç´æ¥å»ä¹°å·ï¼èªè¡æç´¢ã注æä¸è¦è¢«éªï¼ä¸è¬ 5 å 以ä¸å¯ä»¥å ¥æï¼çå°æ 120 ç¾å ç keyï¼è¿ç§å±äºæ¯ç»äºèæä¿¡ç¨å¡ï¼å¯ä»¥éæ¯ 120 ç¾å ï¼åªè½ç¨ä¸ä¸ªæï¼èä¸å®¹æå°å·ãç°å¨æ²¡ç»å¡çè´¦å·éå¶æ¯è¾å¤ï¼å¹¶ä¸å®¹æå°å·ãç»äºå¡ä¹ä¼ä¼å æ£é¤èµ éçä½é¢ã
- è¾å
¥æ¡
- Enteråéï¼å¯å ³éï¼ï¼Shift+Enteræ¢è¡ã
- ç©ºæ ¼ æè / æç´¢ Prompt é¢è®¾ï¼å¯ä»¥æ¨¡ç³å¹é ãææ Prompt å¯ä»¥æ¥ç prompts.md ã
- ä½¿ç¨ ç©ºæ ¼ç©ºæ ¼ æè // æç´¢å ¶å®å¯¹è¯ï¼å¯ä»¥æ¨¡ç³å¹é ï¼æç´¢åå²æ¶æ¯ã
- â å°æè¿çä¸æ¬¡æé®å¡«å°è¾å ¥æ¡éã
- æ¶æ¯ä¸çå¨ä½
- æé®
- ç¼è¾ï¼ä¸è½åå°ç¼è¾ï¼åªè½èªå¨å¡«å°è¾å ¥æ¡éã
- éæ°åçï¼å é¤å½å对è¯ï¼ç¶åèªå¨åéã
- å é¤ï¼å é¤å½å对è¯ã
- åç
- å¤å¶
- éæ°åç
- å é¤ï¼åªå é¤åçã
- æé®
- ç¹å»é¡¶é¨æ é¢æ»å¨å°é¡¶é¨ï¼ç¹å»è¾å ¥æ¡æ»å¨å°åºé¨ã
- url é使ç¨
url?q=ä½ å¥½å
è¿ç§æ¹å¼å¯ä»¥æå¼ç½é¡µç´æ¥åçä½ å¥½å
ï¼å½ä½æç´¢å¼æ使ç¨ãææç对è¯ä¸é½å¯ä»¥ä½¿ç¨ã
设置ä¸å¨ä½
å³è¾¹å个æé®æ¯å¨ä½ï¼ç¹å»å°±å¯ä»¥æ§è¡ã
- ä¼ªè£ è§è²ï¼ä¼ªè£ æ¶ç´æ¥è¿å ¥å¯¹è¯éï¼ä¸ä¼åé请æ±ãå¯ä»¥ä¼ªè£ æç¨æ·æé®ä»¥å GPT åçãåéåèªå¨å ³éä¼ªè£ ã
- å°å½å对è¯çæå¾çï¼æ¹ä¾¿å享ãå¨çµèä¸ç´æ¥å¤å¶å°åªè´´æ¿ï¼ææºä¸æ¯ä¸è½½ã
- å°å½å对è¯çæ Markdownï¼ç´æ¥å¤å¶å°åªè´´æ¿ã
- æ¸ ç©ºå¯¹è¯ã
ç°å¨æ¯æäºå¤ä¸ªå¯¹è¯ï¼ä¹å°±æ¯å¤ä¸ªè天ãä¸åç对è¯å¯ä»¥åç¬è®¾ç½®ï¼ä¹æä¸åçé¾æ¥ï¼æ é¢ï¼å¾æ ãæ以ç°å¨æäºä¸¤ä¸ªè®¾ç½®æé®ï¼ç¬¬ä¸ä¸ªæ¯å ¨å±è®¾ç½®ï¼å¯ä»¥å¡«åå¯ç ï¼API Keyï¼ä¹å¯ä»¥å ³é Enter é®åéï¼è¿æ ·å¨ææºä¸ä¹å¯ä»¥æ¢è¡ã
第äºä¸ªæ¯å¯¹è¯è®¾ç½®ï¼å°½å½å对è¯ææï¼é对ä¸åçç¨éèªè¡è®¾ç½®ãç®å GPT4 ä»ç¶å¤äº waitlist é¶æ®µï¼å¦æ没æèµæ ¼ï¼é£ä¹æ æ³ä½¿ç¨ã
æç»´åæ£ç¨åº¦è¶é«ï¼ChatGPT å°±ä¼ä¹±çï¼å¦ææ¯å¼æ¾æ§çé®é¢ï¼å¯ä»¥éå½è°é«ï¼ä½ä¹ä¸è¦è¶ è¿ 1ãéè¦ç¡®åçæ¡ç建议è°ä¸º 0ã
å¼å¯ è®°å½å¯¹è¯å
容
åï¼å¯¹è¯å·æ°ä¹ä¸ä¼æ¸
空ãå¼å¯è¿ç»å¯¹è¯åï¼æ¯æ¬¡ä¼åéææææä¸ä¸æï¼ä¸å¼å§æè¿ææä¸ä¸æè¿ä¸ªæ¦å¿µã
å½ä½ ç¹å»äºè®¾ç½®æé®ä¹åï¼ä¼åç°å³è¾¹çå¨ä½æé®ä¹åçäºååãæå¼å ¨å±è®¾ç½®åï¼åºç°ç导åºåå¯¼å ¥æé®ãå¯ä»¥å¯¼åºåå¯¼å ¥ææç设置å对è¯ã
èæå¼å¯¹è¯è®¾ç½®ä¹åï¼å¨é¦é¡µï¼ä¹å°±æ¯ä¸»å¯¹è¯éï¼å°±åªæä¸ä¸ªæ°å»ºå¯¹è¯çæé®ï¼å½ä½ æ°å»ºäºå¯¹è¯ãå°±ä¼åºç°å¤å¶å¯¹è¯é¾æ¥ä»¥åå é¤å¯¹è¯çæé®ãå¼å¾æ³¨æçæ¯ï¼ç®åè¿ä¸ªé¾æ¥åªè½ä½ èªå·±è½æå¼ï¼å¹¶æ²¡æä¸ä¼ å°äºç«¯ã
å¨æ°ç对è¯éï¼ä½ å¯ä»¥è®¾ç½®æ é¢ï¼çè³å¯ä»¥è®¾ç½®å¾æ ï¼è¯è¯å¨æ é¢åé¢å ä¸ä¸ª Emojiã
é¨ç½²ä¸ä¸ªä½ èªå·±ç ChatGPT ç½ç«ï¼å è´¹ï¼å½å ç´è¿ï¼
Warning
vercel.app ååå·²ç»è¢«å¢ï¼ä½ vercel æ¬èº«æ²¡æ被å¢ï¼æä»¥ä½ ç»å®èªå·±çååå°±å¯ä»¥äºãå¦æ广æ³å享ï¼ååæ被å¢çé£é©ãå¦ææ¯å½å è´ä¹°çååï¼æ请å»åè¶çé£é©ï¼æ以è¦ä¹å°èå´å å¯ç 使ç¨ï¼è¦ä¹å¨å½å¤è´ä¹°ååãæ´æ¨èå°èå´å 使ç¨ã
å¦æä½ åªéè¦é¨ç½²ä¸ä¸ªä½ èªå·±ç¨çç½ç«ï¼èä¸éè¦å®å¶ï¼é£ä¹ä½ å®å ¨ä¸éè¦å¨æ¬å°è·èµ·æ¥ï¼ä½ å¯ä»¥ç´æ¥ç¹å»ä¸é¢çæé®ï¼ç¶åæç §æ示æä½ï¼ç¶åå¨ Vercel ä¸å¡«å ¥ç¯å¢åéå³å¯ã
ä¸è¿ä¸é¢è¿ç§æ¹å¼ä¸å®¹ææ´æ°ï¼æ好è¿æ¯å
fork æ¬ä»åºï¼ç¶åå¨ Vercel ä¸å¯¼å
¥ä½ èªå·±çä»åºï¼ä¹åè¦æ´æ°å°±å¨ Github éç¹å» Sync fork
å°±å¯ä»¥åæ¥æ´æ°äºã
å¦æä½ éè¦é¨ç½²ç»æ´å¤äººç¨ï¼éè¦ä¿®æ¹ä¸äºä»£ç ï¼é£ä¹ä½ å¯è½éè¦å°ä¸é¢å建çä½ èªå·±çä»åº git clone
å°æ¬å°ãæ¹å®äº git commit & push
å³å¯éæ°é¨ç½²ï¼vscode ä¸ç¹å ä¸å°±å¯ä»¥äºãä¹å¯ä»¥ç¨ vercel ç cliï¼vercel deploy --prod
ã
å¦æä½ éè¦å¨æ¬å°å¼ååè°è¯ï¼æç¹éº»ç¦ï¼
- å级å°
node18
ï¼è¦ç¨å°åççfetch
åreadableStream
ã - API 被å¢äºï¼èªå·±æ³åæ³å¼ä»£çï¼ä¸ç¶è¦æ¥éãå¯ä»¥è®¾ç½® OpenAI ç代ç APIï¼ä¹å¯ä»¥ç´æ¥
vercel deploy
é¨ç½²å° vercel å¼åç¯å¢ä¸è°è¯ã pnpm i
å®è£ ä¾èµãpnpm dev
å¯å¨é¡¹ç®ã
æ´å¤é¨ç½²æ¹æ¡
ç®åæ¬é¡¹ç®é¤ Vercel 以å¤è¿æ¯æ
- Netlifyï¼ç´æ¥å¯¼å ¥å³å¯ã
- CloudFlare Workerï¼ä½¿ç¨ Github Action é¨ç½²ï¼éè¦è®¾ç½®
CF_API_TOKEN
ã - Railwayï¼ç´æ¥å¯¼å ¥å³å¯ã
ä¸å»ºè®®ä½¿ç¨é¤ Vercel 以å¤çé¨ç½²æ¹æ¡ã é¤äºæ¯è¾æ ¢ä»¥å¤ï¼æ´éè¦çæ¯åªæ Vercel æ¯æ设置 Edge Function æå¡å¨çå°åºï¼å ¶ä»å¹³å°ä¼èªå¨ä½¿ç¨è·ç¦»æè¿çæå¡å¨ï¼æå¯è½æ¯ OpenAI ä¸æ¯æçå°åºï¼ä»è导è´å°å·ã
ç¯å¢åé
以
CLIENT_
å¼å¤´çåéä¼æ´é²ç»å端ï¼è¯·ä¸è¦å¡«åææä¿¡æ¯ã
ç¯å¢åé | 说æ | é»è®¤å¼ |
---|---|---|
OPENAI_API_KEY | OpenAI API Keyï¼å¯ä»¥å¡«åå¤ä¸ªï¼ç¨ | æè æ¢è¡ éå¼ï¼éæºè°ç¨ãæ好æ¯å¤å¡«å 个ï¼API æ并åä¸çéå¶ãå¦æç¨æ·ä¸å¡«èªå·±ç keyï¼é£ä¹å°±ä¼ä½¿ç¨ä½ ç keyã | æ |
OPENAI_API_BASE_URL | æ¬å°å¼åæ¶å¯ä»¥å¡«å OpenAI ç代çæå¡å¨ï¼ä½æ¯ Vercel ä¸éè¦ã并ä¸ä¸å»ºè®®ç产é¶æ®µä½¿ç¨ã | api.openai.com |
NO_GFW | 表示å¯ä»¥æå¡å¨å¯ä»¥ç´è¿ï¼ä¸éè¦OPENAI_API_BASE_URL ï¼å°±ç®è®¾ç½®äºä¹ä¸ä½¿ç¨ã | false |
TIMEOUT | OpenAI API è¶ æ¶ãåä½ä¸ºæ¯«ç§ï¼Vercel Edge Function æå¤§è¶ æ¶ä¸º 30000ã | 30000 |
PASSWORD | ç½ç«å¯ç | æ |
CLIENT_DEFAULT_MESSAGE | é»è®¤æç¤ºä¿¡æ¯ | - xx xx |
CLIENT_GLOBAL_SETTINGS | é»è®¤å ¨å±è®¾ç½® | {"APIKey":"","password":"","enterToSend":true} |
CLIENT_SESSION_SETTINGS | é»è®¤å¯¹è¯è®¾ç½®ï¼å¯¹è¯è®¾ç½®å¨æ¯ä¸ªå¯¹è¯ä¸é½æ¯ç¬ç«çã | {"title":"","saveSession":true,"APITemperature":0.6,"continuousDialogue":true,"model":"gpt-3.5"} |
CLIENT_MAX_INPUT_TOKENS | ææä¸ä¸æ+è¾å ¥ç token æ°ãOpenAI ä¸å模åçæ大 token ä¸ä¸æ ·ï¼ä»·æ ¼ä¹ä¸åï¼å¯ä»¥åå«è®¾ç½®ãå¹¶ä¸ OpenAI ä¼ç»è®¡è¾å ¥+è¾åºä¹åï¼ä½æ们è¿éåªéå¶è¾å ¥ã | {"gpt-3.5":16000,"gpt-4":32000} |
æ两ç§è®¾ç½®æ¹å¼
-
å°
.env.example
æ件修æ¹ä¸º.env
ï¼å¨.env
ä¸è®¾ç½®ã -
Vercel ä¸è®¾ç½®
Environment Variables
ãå°½é使ç¨è¿ç§æ¹å¼ï¼æ¯è¾æ¹ä¾¿ãä¼å¨ä¸æ¬¡é¨ç½²æ¶çæã
é»è®¤å ¨å±è®¾ç½®
è®°å¾å é¤æ³¨éï¼æè ç´æ¥å¤å¶ä¸é¢è¡¨æ ¼éçã
{
"APIKey": "", // é»è®¤å¡«åç keyï¼ä¸éè¦å¡«åï¼å¦åå
¶ä»äººçå¾å°ã
"password": "", // é»è®¤å¡«åçå¯ç ï¼ä¸éè¦å¡«åï¼å¦åå
¶ä»äººçå¾å°ã
"enterToSend": true // å车é®åéæ¶æ¯
}
é»è®¤å¯¹è¯è®¾ç½®
对è¯è®¾ç½®å¨æ¯ä¸ªå¯¹è¯ä¸é½æ¯ç¬ç«çãè®°å¾å é¤æ³¨éï¼æè ç´æ¥å¤å¶ä¸é¢è¡¨æ ¼éçã
{
"title": "", // 对è¯æ é¢ï¼ä¸éè¦å¡«å
"saveSession": true, // è®°å½å½å对è¯å
容ï¼å·æ°ä¸ä¼ä¸¢å¤±ãå
³éåä»ç¶ä¼è®°å½éå®ç对è¯ã
"APITemperature": 0.6, // 0-2ï¼æç»´åæ£ç¨åº¦ï¼è¶é« ChatGPT æç»´å°±è¶åæ£ï¼å¼å§ä¹±çï¼çè³ä¼ä¹±ç ï¼å»ºè®®å°äº 1 ã
"continuousDialogue": true, // å¼å¯è¿ç»å¯¹è¯ï¼æ¯æ¬¡é½éè¦å°ä¸ä¸æä¼ ç» APIã
"model": "gpt-4o-mini" // 模å
}
æäº¤ä½ ç Prompts
- Fork æ¬é¡¹ç®ã
- ä¿®æ¹
prompts.md
ã - Pull Request å³å¯ã
å¦æä½ ä¸æè¿ä¸ªæä½ï¼ä¹å¯ä»¥ç´æ¥å¨ Issues æäº¤ä½ ç Promptsãç®å大é¨å Prompts æ¥èªäº awesome-chatgpt-prompts-zhï¼å½ç¶ï¼è¿ä¸ªä»åºå¤§å¤æ°ä¹æ¯ç¿»è¯ç awesome-chatgpt-promptsï¼ä¸å¹¶æè°¢ã
è¦æ±
- æéè¦è¾å ¥çå 容æ¾å¨æåï¼å¯ä»¥æ示 ChatGPT å¼å§è¾å ¥äºï¼æ¯å¦ âæç第ä¸å¥è¯æ¯ï¼âã
- å°½å¯è½å»ä¼åå·²æç Promptsï¼èä¸æ¯éå¤æ·»å ã
- æ·»å å°ç»å°¾ï¼æä¼å®ææ´çã
èµèµ
å¦ææ¬é¡¹ç®å¯¹ä½ ææ帮å©ï¼å¯ä»¥ç»å°ç«ä¹°ç¹é¶é£ï¼ä½ä¸æ¥åä»»ä½ä»è´¹åè½è¯·æ±ã
License
Top Related Projects
AI agent stdlib that works with any LLM and TypeScript AI SDK.
🔮 ChatGPT Desktop Application (Mac, Windows and Linux)
Use ChatGPT On Wechat via wechaty
用 Express 和 Vue3 搭建的 ChatGPT 演示网页
A cross-platform ChatGPT/Gemini UI (Web / PWA / Linux / Win / MacOS). 一键拥有你自己的跨平台 ChatGPT/Gemini 应用。
Convert designs to code with AI
Introducing Visual Copilot: A new AI model to turn Figma designs to high quality code using your components.
Try Visual Copilot