askrella / whatsapp-chatgpt

ChatGPT + DALL-E + WhatsApp = AI Assistant :rocket: :robot:

GPT + DALL-E + WhatsApp = AI Assistant 🚀

This WhatsApp bot uses OpenAI's GPT and DALL-E 2 to respond to user inputs.

You can talk to the bot in voice messages, the bot will transcribe and respond. 🤖

Requirements

• Node.js (18 or newer)
• A recent version of npm
• An OpenAI API key
• A WhatsApp account

Documentation

In the documentation you can find more information about how to install, configure and use this bot.

➡️ https://askrella.github.io/whatsapp-chatgpt

Disclaimer

The operations performed by this bot are not free. You will be charged by OpenAI for each request you make.

This bot uses Puppeteer to run a real instance of Whatsapp Web to avoid getting blocked.

NOTE: We can't guarantee that you won't be blocked using this method, although it does work. WhatsApp does not allow bots or unofficial clients on its platform, so this should not be considered completely safe.

Contributors

Used libraries

• https://github.com/transitive-bullshit/chatgpt-api
• https://github.com/pedroslopez/whatsapp-web.js
• https://github.com/askrella/speech-rest-api

whatsapp-chatgpt's Issues

I would like to propose `dalle-config` instead of `config`, we may have the same for gpt or others ai assistants.

          I would like to propose `dalle-config` instead of `config`,  we may have the same for gpt or others ai assistants.

Originally posted by @andrewssobral in #62 (comment)

Exception

Exception:

file:///E:/whatsapp-chatgpt-master/whatsapp-chatgpt-master/node_modules/chatgpt/build/index.js:132
throw new Error("Invalid environment; fetch is not defined");
^

Error: Invalid environment; fetch is not defined
at new ChatGPTAPI (file:///E:/whatsapp-chatgpt-master/whatsapp-chatgpt-master/node_modules/chatgpt/src/chatgpt-api.ts:152:13)
^
Error: Invalid environment; fetch is not defined
at new ChatGPTAPI (file:///E:/whatsapp-chatgpt-master/whatsapp-chatgpt-master/node_modules/chatgpt/src/chatgpt-api.ts:152:13) at E:\whatsapp-chatgpt-master\whatsapp-chatgpt-master\src\index.ts:10:13

ChatGPTAPI error 400

How to solve ChatGPTAPI error 400 ?

add a pipeline and some unit tests ?

any suggestions on what to use? Mocha? Tape?

Fail to run "npm run start"

I got the error below, anyone can help?

[email protected] start C:\Users\kuang\Desktop\WhatsApp_ChatGPT
npx vite-node src/index.ts

npm WARN config cache-max This option has been deprecated in favor of --prefer-online
npm WARN config cache-min This option has been deprecated in favor of --prefer-offline.
npm WARN config optional Use --omit=optional to exclude optional dependencies, or
npm WARN config --include=optional to include them.
npm WARN config
npm WARN config Default value does install optional deps unless otherwise omitted.
npm WARN config shrinkwrap Use the --package-lock setting instead.
npm WARN config tmp This setting is no longer used. npm stores temporary files in a special
npm WARN config location in the cache, and they are managed by
npm WARN config cacache.
npm ERR! code ENOLOCAL
npm ERR! Could not install from "Files\nodejs\node_cache_npx\12112" as it does not contain a package.json file.

npm ERR! A complete log of this run can be found in:
npm ERR! C:\Program Files\nodejs\node_cache_logs\2023-02-08T15_47_41_230Z-debug.log
Install for [ 'vite-node@latest' ] failed with code 1
npm ERR! code ELIFECYCLE
npm ERR! errno 1
npm ERR! [email protected] start: npx vite-node src/index.ts
npm ERR! Exit status 1
npm ERR!
npm ERR! Failed at the [email protected] start script.
npm ERR! This is probably not a problem with npm. There is likely additional logging output above.

npm ERR! A complete log of this run can be found in:
npm ERR! C:\Program Files\nodejs\node_cache_logs\2023-02-08T15_47_41_327Z-debug.log

Dalle not showing pics

Hi there,
the !dalle command works because it answers to me but the pics i try to generate don't show up. Just a description of the actual photo i requested comes out. What could be wrong?

Chatgpt non showing up among my others chats.

Hi there,
it seems that what i have done so far is working because i get the message " [Whatsapp ChatGpt] Client is Ready! " and all the texts that i receive are being written in my terminal and the bot is replying on my behalf but the actual chat is not showing up, so i can't ask question. What is it that i could've done wrong? Also, i do i stop my bot to answer when the whatsapp session is active? (for now i just interrupted the session. Let me know.
Thank you

Error: Evaluation failed: TypeError: Cannot read properties of undefined (reading 'LegacyPhoneFeatures')

heaving the following error with it :

C:\wamp64\www\whatsapp-chatgpt\node_modules\whatsapp-web.js\node_modules\puppeteer\lib\cjs\puppeteer\common\ExecutionContext.js:221
throw new Error('Evaluation failed: ' + helper_js_1.helper.getExceptionMessage(exceptionDetails));
^

Error: Evaluation failed: TypeError: Cannot read properties of undefined (reading 'LegacyPhoneFeatures')
at puppeteer_evaluation_script:13:76
at ExecutionContext._evaluateInternal (C:\wamp64\www\whatsapp-chatgpt\node_modules\whatsapp-web.js\node_modules\puppeteer\src\common\ExecutionContext.ts:273:13)
at processTicksAndRejections (node:internal/process/task_queues:95:5)
at ExecutionContext.evaluate (C:\wamp64\www\whatsapp-chatgpt\node_modules\whatsapp-web.js\node_modules\puppeteer\src\common\ExecutionContext.ts:140:12)
at Client.initialize (C:\wamp64\www\whatsapp-chatgpt\node_modules\whatsapp-web.js\src\Client.js:267:9)

Node.js v18.13.0

Voice message transcription feature

Already working on feature/whisper branch.

• We need some documentation in docs/whisper.md on how to install whisper.
  • Windows, Linux, Mac
• Needs better code structure, i just drafted it
• Needs more testing

[Whatsapp ChatGPT] Failed to authenticate with the ChatGPT API: No element found for selector: button[type="submit"]

I Followed the instruction and got following issue after npm run start

[Whatsapp ChatGPT] Failed to authenticate with the ChatGPT API: No element found for selector: button[type="submit"]

It went till email and password screen but after entering password it stopped the process with above error.

Automatically deciding what model to use

Currently, if you don't provide a prefix, the assistant will only use GPT. However, it is possible to analyze the user's request and decide which model to use.

Alternatively, maybe even sending a (hidden) request to GPT and letting it decide if it should handle the request or if DALLE should could be a viable solution. Obviously, (regardless of approach) this could be a relatively large feature, especially if you would want to increase the accuracy of these 'guesses' by providing additional context to whatever model/regex/whatever such as previous messages.

This would make the assistant feel more complete as you can have a more natural 'conversation' with it.

Sorry if this is a duplicate suggestion, I could not find one similar to this.

.ENV file

I cannot find the ENV file in order to input the OpenAI user and password

Available as a Docker Image?

It would be amazing if there were a Docker image and docker-compose file available for easy set for us less tech savvy!

Automation can't proceed to login due to recaptcha

As instructed the chrome opens and enters the email but due to reCAPTCHA the code won't proceed

Why not using OPENAI_API_KEY ? (instead of EMAIL + PASSWORD)

This is dope **BUT**

Please add a list of the things to avoid/not do in the read me file.

What happen if I spam questions ? (Should I wait a little between each time I use !gpt ?)
Can my account get banned for using this code ?

Immediately after seeing the read me file I got the idea of making a group chat with my cs buddies where they can all use my chatgpt plus account for a dollar per month (or less based on the number of individuals within the group, so we can all enjoy chatgpt plus practically for free).
I suppose chatgpt wouldn't want something like that to happen and would do their best to prevent and forbid such accounts.

Feedback Mechanism

Can we have a Feedback Mechanism for getting the user feedback if the response from chat GPT or Dall E is what the user is looking for? Right now Chat GPT has a Like or Dislike button to rate the generated text.

i get this error when scanning the QR code

i don't have any big experience in coding or anything, is there a fix?

.env file not found

i downloaded the Zip file und it has no .env file included so that i can put my E-email and password.
do i need need do create it Manuelly ? and in which path ? i am installing it on MacOs

i get the following output:

ChatGPTError: ChatGPT invalid email
at new ChatGPTAPIBrowser (file:///Users/cdmi/Documents/Projects/whatsapp-chatgpt-master/node_modules/chatgpt/src/chatgpt-api-browser.ts:102:21)
at /Users/cdmi/Documents/Projects/whatsapp-chatgpt-master/src/index.ts:17:13
at ViteNodeRunner.directRequest (file:///Users/cdmi/.npm/_npx/f2342a4b64a2bc92/node_modules/vite-node/dist/client.mjs:308:5)
at ViteNodeRunner.cachedRequest (file:///Users/cdmi/.npm/_npx/f2342a4b64a2bc92/node_modules/vite-node/dist/client.mjs:156:14)
at ViteNodeRunner.executeFile (file:///Users/cdmi/.npm/_npx/f2342a4b64a2bc92/node_modules/vite-node/dist/client.mjs:133:12)
at CAC.run (file:///Users/cdmi/.npm/_npx/f2342a4b64a2bc92/node_modules/vite-node/dist/cli.mjs:64:5) {
statusCode: 401
}

How can i run this on a VPS ?

When i try to run this on a VPS i get this error:

            reject(new Error([
                   ^

Error: Failed to launch the browser process!
/home/siu/whatsapp-chatgpt/node_modules/whatsapp-web.js/node_modules/puppeteer/.local-chromium/linux-982053/chrome-linux/chrome: error while loading shared libraries: libatk-1.0.so.0: cannot open shared object file: No such file or directory


TROUBLESHOOTING: https://github.com/puppeteer/puppeteer/blob/main/docs/troubleshooting.md

    at onClose (/home/siu/whatsapp-chatgpt/node_modules/whatsapp-web.js/node_modules/puppeteer/src/node/BrowserRunner.ts:268:9)
    at Interface.<anonymous> (/home/siu/whatsapp-chatgpt/node_modules/whatsapp-web.js/node_modules/puppeteer/src/node/BrowserRunner.ts:254:50)
    at Interface.emit (node:events:524:35)
    at Interface.close (node:internal/readline/interface:534:10)
    at Socket.onend (node:internal/readline/interface:260:10)
    at Socket.emit (node:events:524:35)
    at endReadableNT (node:internal/streams/readable:1359:12)
    at processTicksAndRejections (node:internal/process/task_queues:82:21)

Node.js v19.6.1

Seems like a problem with puppeteer not being able to open a browser instance...

Could you help me run this on a VPS with no UI?

Not able to see WhatsApp code

I don't get whatsapp code to scan . please help @navopw thanks in advance

(node:20828) ExperimentalWarning: The Fetch API is an experimental feature. This feature could change at any time
(Use node --trace-warnings ... to show where the warning was created)
node:internal/errors:484
ErrorCaptureStackTrace(err);
^

Error: spawn EPERM
at ChildProcess.spawn (node:internal/child_process:413:11)
at Object.spawn (node:child_process:743:9)
at BrowserRunner.start (C:\Users\Bhavya.Damani\OneDrive - DynaTech Systems Pvt. Ltd\Desktop\whatsapp-chatgpt-master\node_modules\whatsapp-web.js\node_modules\puppeteer\src\node\BrowserRunner.ts:89:30)
at ChromeLauncher.launch (C:\Users\Bhavya.Damani\OneDrive - DynaTech Systems Pvt. Ltd\Desktop\whatsapp-chatgpt-master\node_modules\whatsapp-web.js\node_modules\puppeteer\src\node\Launcher.ts:162:12)
at Client.initialize (C:\Users\Bhavya.Damani\OneDrive - DynaTech Systems Pvt. Ltd\Desktop\whatsapp-chatgpt-master\node_modules\whatsapp-web.js\src\Client.js:100:23) {
errno: -4048,
code: 'EPERM',
syscall: 'spawn'

Is there a way to send yourself questions (not just receive)?

I would like a way to interact with the bot for myself, not just set this up for other people to use, how is that done, do we need to setup whatsapp for another phone number? Or is there a way to set up whatsapp for business to make that work?

Only a question: What is meaning with the disclaimer

The operations performed by this bot are not free. You will be charged by OpenAI for each request you make

Hello,

i do not understand this. If i have a free open ai account, do the bot not work?
Thanks for clarification ...

Best Regards

invalid email

HI,

Im getting an invalid email when starting. env is created and contains the credentials with the valid email for the openai account.

Error after Inactivity

I installed and set up the bot correctly. It also works perfectly. Now I wanted to keep the bot accessible 24/7 and put it on my server. After a period of inactivity, the bot stops responding and throws an error. Does anyone have any idea how I can get around this so that I can message the bot 24/7?

(ChatGPTAPI error 429)

Everything works great. I got a reply for the first message but for the rest i get the following error as a reply.

"An error occured, please contact the administrator. (ChatGPTAPI error 429)"

Any fix would be appreciated..

Providing Custom Context

It would be nice if we let users add their own context.

Example

const custom_context = `
 Context:
 RandomsTS is a free and open-source framework for building restful APIS 
 using typescript and class-based object-oriented programming.
 Key Features: file-base routing, Hot-reloading, Type safety, Intuitive APIs.
 
 NOTE! Only answer from the given context
 question: 
 <add user_message here>
`

const canUseCustomContext = messageStr.includes("RandomsTS");
if (canUseCustomContext) 
{
  // add custom_context before user message
  await handleMessageGPT(message, custom_context+prompt)
}

Results

 message:- what is RandomsTS
 
 reply:- RandomsTS is a free and open-source framework for building restful APIS 
 using typescript and class-based object-oriented programming.

Not responding

When I try to say something it doesn't respond. I tried to message myself. Can only other people whatsapp me?

Test

Exceeded Current Quota

I started the project locally, and successfully got it set up. But when I try to use gpt, here's what I get

I don't even use ChatGPT so often that it should show me the error Exceeded Current Quota.

Kindly help me out

Regards
Sam Varghese

Failed to authenticate with the ChatGPT API ,ChatGPT failed to refresh auth token

My concern about this is not very clear to be fair! What is the problem with these API Token?

Custom Context (`!context <name>`)

Define a json configuration file that has a list of actors (e.g. doctor, humorist, etc).

context-config.json or something similar

[
    {
        name: "doctor",
        prompt: "Act as a doctor, give medical advise"
    },
    {
        name: "humorist",
        prompt: "Act as a humorist"
    }
]

Add the environment variable CONTEXT which defines the actor/context for all newly created conversations.

Make it possible to switch between different context/actors with !context <name>

Example code for promptPrefix:

const res = await api.sendMessage('what is the answer to the universe?', {
  promptPrefix: `Act as ...`
})

Credits go to: @andrewssobral

loading Spinner while loading chat

Problem

When the user scans QR after this what's the app takes some time to load the chat.

Solution

It would be nice to add a loading spinner while the chat is being loaded for a better experience.

Example

client.on('loading_screen', (percent, message) => {
        // show spinner here
});

Where is the .env file?

sorry, im having trouble finding the .env file.
thank you

npm WARN deprecated [email protected]: < 18.1.0 is no longer supported

انا رضا ديكور

          انا  رضا ديكور

Originally posted by @Re1da in #43 (comment)

A way to store the session or login details.

When we turn off the program or it got interrupted by any other means, we have to scan the code again, instead if it saves the login details in a file, it will be good.

can't rename to .env in certain versions of windows.

We need to resolve this issue for windows users.

Error 403

Authentication was successful but its not working.

Whatsapp Reply:

An error occured, please contact the administrator. (ChatGPTAPI error 403)

Terminal:

[Whatsapp ChatGPT] Client is ready!
[Whatsapp ChatGPT] Received prompt from <mynumber>@c.us: hi

ازكwhen you will merge it?

          when you will merge it?

Originally posted by @Zain-ul-din in #43 (comment)

le navigateur s'ouvre puis se referme

Bonjour,

J'ai suivi toutes les étapes sur Mac OS. Le programme fonctionne bien, mais la fenêtre du navigateur s'ouvre, puis se ferme directement. Le terminal me montre ceci :
ChatGPTError : Jeton de session ChatGPT non valide
Code d'état : 401

Error after inactivity

Is it possible to run the bot permanently on a server? I set up the bot correctly and put it on my server. First the bot works, but after 20 minutes without any questions (requests), an error occurs when I ask the bot a question. Probably the AI ​​just stops working after a period of inactivity. Now I would like to bypass that so I have a 24/7 whtasapp chatbot. Does somebody has any idea?

Get more seats for this project since it's organizational open source

Where do i add the .env file?

HY

Error 429

Please translate it to English. Thank you for your great work in advance! Once I send a request with !gpt, I receive the response

"An error occurred, please contact the administrator. (ChatGPTAPI error 429)".

I immediately asked ChatGPT about the error message and received the response that too many requests were made. Well, the openai account is new and only four chats have been started with it. Any ideas?

[Whatsapp ChatGPT] Client is ready! [Whatsapp ChatGPT] Received prompt from [email protected]: Wie viele Zeitzonen gibt es? An error occured ChatGPTError: ChatGPTAPI error 429 at ChatGPTAPIBrowser.sendMessage (file:///C:/Users/adm/whatsapp-chatgpt/node_modules/chatgpt/src/chatgpt-api-browser.ts:563:23) at processTicksAndRejections (node:internal/process/task_queues:95:5) at handleMessage (C:\Users\adm\whatsapp-chatgpt\src\index.ts:48:22) at Client.<anonymous> (C:\Users\adm\whatsapp-chatgpt\src\index.ts:36:9) { statusCode: 429, statusText: '' }

Doesn't run on Raspberry Pi 4

pi@pi:~/whatsapp-chatgpt $ npm run start

[email protected] start
npx vite-node src/index.ts

┌ Whatsapp ChatGPT & DALL-E
│
◇ ────────────────────────────────────────────────────────────────────────────────────────────────╮
│ │
│ A Whatsapp bot that uses OpenAI's ChatGPT and DALL-E to generate text and images from a prompt. │
│ │
├───────────────────────────────────────────────────────────────────────────────────────────────────╯
│
○ Starting
/home/pi/whatsapp-chatgpt/node_modules/whatsapp-web.js/node_modules/puppeteer/lib/cjs/puppeteer/node/BrowserRunner.js:241
reject(new Error([
^

Error: Failed to launch the browser process!
/home/pi/whatsapp-chatgpt/node_modules/whatsapp-web.js/node_modules/puppeteer/.local-chromium/linux-982053/chrome-linux/chrome: 16: Syntax error: ")" unexpected

TROUBLESHOOTING: https://github.com/puppeteer/puppeteer/blob/main/docs/troubleshooting.md

at onClose (/home/pi/whatsapp-chatgpt/node_modules/whatsapp-web.js/node_modules/puppeteer/src/node/BrowserRunner.ts:268:9)
at Interface.<anonymous> (/home/pi/whatsapp-chatgpt/node_modules/whatsapp-web.js/node_modules/puppeteer/src/node/BrowserRunner.ts:254:50)
at Interface.emit (node:events:525:35)
at Interface.close (node:internal/readline/interface:533:10)
at Socket.onend (node:internal/readline/interface:259:10)
at Socket.emit (node:events:525:35)
at endReadableNT (node:internal/streams/readable:1359:12)
at processTicksAndRejections (node:internal/process/task_queues:82:21)

Node.js v18.14.1

No Answer

error creating qr code

Error: Cannot find module 'qrcode-terminal'
Require stack:

• C:\Users\Ron Rull\Downloads\Compressed\whatsapp-chatgpt-master\src\index.ts
  at Function.Module._resolveFilename (node:internal/modules/cjs/loader:1039:15)
  at Function.Module._load (node:internal/modules/cjs/loader:885:27)
  at Module.require (node:internal/modules/cjs/loader:1105:19)
  at require (node:internal/modules/cjs/helpers:103:18)
  at C:\Users\Ron Rull\Downloads\Compressed\whatsapp-chatgpt-master\src\index.ts:2:16
  at ViteNodeRunner.directRequest (file:///C:/Users/Ron%20Rull/AppData/Local/npm-cache/_npx/f2342a4b64a2bc92/node_modules/vite-node/dist/client.mjs:308:11)
  at processTicksAndRejections (node:internal/process/task_queues:95:5)
  at ViteNodeRunner.cachedRequest (file:///C:/Users/Ron%20Rull/AppData/Local/npm-cache/_npx/f2342a4b64a2bc92/node_modules/vite-node/dist/client.mjs:156:14)
  at ViteNodeRunner.executeFile (file:///C:/Users/Ron%20Rull/AppData/Local/npm-cache/_npx/f2342a4b64a2bc92/node_modules/vite-node/dist/client.mjs:133:12)
  at CAC.run (file:///C:/Users/Ron%20Rull/AppData/Local/npm-cache/_npx/f2342a4b64a2bc92/node_modules/vite-node/dist/cli.mjs:64:5) {
  code: 'MODULE_NOT_FOUND',
  requireStack: [
  'C:\Users\Ron Rull\Downloads\Compressed\whatsapp-chatgpt-master\src\index.ts'
  ]
  }

Issue when i authenticate using google in chat.openai.com

When i authenticate using google in chatgpt, it seems like it does not recognize the login and wont redirect me to whatsapp. It stays stuck or something wont process the next step, not sure.