The BEST ESP32 project based on ChatGPT | IOT Projects | NextPCB
HTML-код
- Опубликовано: 7 фев 2025
- Ready to take your ESP32 projects to the next level? In this video, we'll show you how to create your own ChatGPT voice assistant using an ESP32 board. This DIY tutorial is perfect for anyone looking to incorporate artificial intelligence into their IOT projects. Watch now to learn how to make your own ChatGPT voice assistant with ESP32!
⚡️⚡️ You can also start learning IOT from scratch with my Course "Master of IOT". Enroll Now - techiesms.grap...
⚡️⚡️ Order all components required for making this project using this single link - techiesms.com/...
You can support me🙏🏻 and also get benefitted on Patreon :- / techiesms
__________________________________________________________
👨🏻💻 NextPCB Links 👨🏻💻
This video is sponsored by NextPCB - the Reliable Multilayer PCB Manufacturer.
www.nextpcb.com
Free international shipping for Authentic Electronic Components:
www.hqonline.com/
Enjoy free PCB assembly for 5 boards:
www.nextpcb.co...
DFM free online PCB Gerber viewer:
www.nextpcb.co...
__________________________________________________________
👨🏻💻 Important Links 👨🏻💻
For OpenAI API - platform.opena...
For Google Cloud API - cloud.google.com/
Postman App Download - www.postman.co...
Codes & Sch Used in this project - github.com/tec...
ESP32 ChatGPT Video Series - • Running ChatGPT on ESP...
__________________________________________________________
Components used in this project 💡🔌
⚡️⚡️ Order all components required for making this project using this single link - techiesms.com/...
Or if you want to order only specific components then here is the list and link,
ESP32 Dev board - www.techiesms....
I2S Mic - www.techiesms....
I2S Audio Amplifier - www.techiesms....
Speaker - www.techiesms....
IR Sensor Module - www.techiesms....
TP4056 Battery Charging Module - www.techiesms....
Battery - www.techiesms....
__________________________________________________________
⏱ Time Stamps ⏱
__________________________________________________________
🎥 🎙 My Video Gear 🎙🎥
Camera - Canon 750d
Lens - 50mm f/1.8 STM Lens (amzn.to/32gqnsi), 24mm f/2.6
Key Light - Godox SL60W (amzn.to/3522mHe)
Fill Light/Back Light - Osaka Lightning Setup (amzn.to/2I5cdDi)
Tripod - amzn.to/3uNhC4H
Microphone - Zoom H1N(amzn.to/3Osh12A)
Editing - Final Cut Pro X on MacBook Pro ( amzn.to/3vHcSyJ )
__________________________________________________________
Get connected with techiesms everywhere..
Subscribe Main Channel :- bit.ly/techiesm...
Subscribe Shorts Channel :- bit.ly/techies...
Facebook :- / techiesms
Twitter :- / imtechiesms
Instagram :- / techiesms
Website :- www.techiesms.com
Telegram - @techiesms
__________________________________________________________
techiesms
explore | learn | share
If we ever make version 3 of this project, what will be your suggestions about adding/removing things in this project? Share your thoughts..
Make it offline. Even tho it seems a little hard
Great work anyways. As always I will keep buying products from you
The thing is ultimately it isn't free
you can do the same thing (generative language) for Gemini Pro and I have a good idea, you can make the project more beautiful by using OLED screen (SSD1306) and Wowki Animator.
Because Gemini API is free and I have another idea:2. ESP (I mean the ESP32 module that controls the speaker) should be in the box, but
1. Let the esp (I mean the esp32 module that controls the microphone) be on a remote control and record the sound when you click the button.
As suggested below: use a wakeup word and create the possibility to 1. run an Internet radio 2. use ChatGPT for question/answer, 3 Perform an action with Servomotor. Integrate also the board DY-HV20 for predefined responses.
Then this project will become a very good foundation for a ROBOT.
Anyway: Sachin congrats: version 2 works!!
Custom PCB design isn't there on the description Can u please upload that
its look amazing ... this time your talking clearly and slowly thanks for this video last video i got so many errors. hope this time its work... i will update once its works...Please keep doing this kind of videos...
I try my best to minimise the errors
So do try it out and let me know if you still get any error
Bro did that work
@@amalnp3934 not working bro
just seen i made your 123.000's youtube abo lol :D, congratulation techiesms :) .. go on with this great work !, locking forward v.3
Thank you ✌️
Is there a specific reason you chose to have two different boards for two different functions? Can’t both of them be uploaded into one?
Unfortunately, ChatGPT API isnt cheap. You incur cost for both input and output , so a device running chatgpt as the interface will quickly run out of credits. A better alternative is to use Hugging chat which is free but it only has python library.
Thanks for the suggestion
Bro help me about Google api is there alternative
You can just scrape the library api and use it in c code instead of the openai api
@@snopz can you please explain it in little more detail if possible or provide with the some links please
I've tried multiple options , from llama based models , home start etc and a few local LLMs all native on my GPU but none of them compare to GPT APi, it is expensive but I've set my spend to £6 a month and use the cheapest 3.5 API and it's just unrivaled, I used it integrated into home assistant and no other LLM comes close , which is unfortunate for the time being
Thank you so much for this latest project.
Hats off for your contribution to the hobbyist community. ❤❤
Hello Techiesms!
The project is working great - great job!
The problem is low processor memory. only 2 seconds for the Questions task is too little. Longer responses are skipped because the receiving buffer is too small. I have a question whether it is possible to modify the program for ESP32 VROVER to extend the Voice Recorder (5 sec) and the receiving buffer for Text 2 Speech (Max Tokens 100).
Then the device would be perfect!
Warm greetings from Bavaria!
Roman
Hello did you make the project as I wanted some tips
Hi techiesms,
your project is awesome .. can't wait for next steps. I will rebuild it for sure, doing some adjustments and enjoy ;). Any ideas for battery mod would be best anyhow. To your questions on eof videe (ideas for next steps), here my personal coins:
1. stay with ESP32, not using Raspberry Pi, reasons e.g: (a) we can utilize the existing code, no Python needed, (b) no Pi (Unix OS) overhead needed, keep it small and beauty, and (c) much easier for battery powering .. just my coin (i jumped once to Pi, did not like it .. then went back to my sweet home ESP32 & Teensy ;)
2. solving the input length issue would be most important (i feel 2 or 3 seconds are too short for nice chats)
3. yes, solving the charging issue, would be awesome to have this beauty box totally portable !
4. later on maybe a 'wake-up-key' phrase (just optionally, i will stay with a simple capacitive touch button for now)
just my coins, I'm exited next steps ! .. cool work !
The audio.h Library shows an error like
In file included from c:\Users\sreerama\Documents\Arduino\libraries\ESP32-audioI2S-master\src\Audio.cpp:15:0:
c:\Users\sreerama\Documents\Arduino\libraries\ESP32-audioI2S-master\src\opus_decoder/opus_decoder.h:7:20: fatal error: memory.h: No such file or directory
compilation terminated.
exit status 1
Compilation error: exit status 1
Same error did you find the solution...? Please help 🙏
@amalnp3934 did you find the solution of this error, please help
Download that directory or library
@@saidevdhalfrom where?
same to same bro
Hello! Can you provide the design of the PCB that you used?
how to get the pcb ??
Can you make a video on how to connect the components properly?
Great project and why can not we use single DevKit for this?
You could use one single esp32 and link it to esphome and make a pipeline for the chat gpt and the tts and stt (tts and stt are free and open source in home assistsnt)
And also you will be able to use wake word instead of the distance sensor.
Bro help me about Google api is there alternative
Bro help me about Google api is there alternative
What free tts and stt you mean? 😮
@@antonpegov2745piper(tts) and whisper(stt) they are both free and open source and included in home assistant as integrations
Where do I get the gerber file of the PCB?
use DC to DC converter and for the trigger use an accelerometer like lis2dw12 which has a power consumption of around 1 microamp, put deep sleep both esp32, and when someone taps on your project or shakes it it gives a trigger, same as your IR sensor, and both esp32 wakeup and ready for normal operations there same delay when wakeups like connect to the router and other process so indicate with light when the system is ready after no voice listening for certain period, go back to sleep overall power consumption should be around 20 uam ,if Load Switch IC use where ever needed and your low power consumption device is ready is may able to run 1 month on 1000 mah battery
Great as always. Really appreciate your great effort
Thank you so much 😀
@@techiesmsBro help me about Google api is there alternative
bro techsims can I make it work without pcb
through jumper wires
You are great, you can do the same thing (generative language) for Gemini Pro and I have a good idea, you can make the project more beautiful by using OLED screen (SSD1306) and Wowki Animator.
For example, when receiving a response, let the OLED screen perform a certain animation, let a logo appear when it is started, etc.
@techiesms can i use Gemini api instead of chat gpt api ?
Bro can you send the link of the PCB design you have build it
Hello can you provide the link of the pcb that has used
How can i get youyr PCB
Hi techiesms .. let us know: could you solve the 2 sec record length issue meanwhile ? I would love to record e.g.5-10 secs. ANY idea so much welcome - thank you!
Great, thanks for share. Love u.
How much does it cost to buy one from you including the schematics and code?
Isn't there free api instead of using Google cloud and OpenAI chatgpy
Code is free idk about schematics and iots are 1500 inr
Hello Sir, can I use the schematic for gemini esp32?
If Possible can We Change The Voice note Of The Assistant with The Help Of Other AI Softwares Like Eleven's lab To Change The Voice Note Of The Output?
can you tell the some advantages of this project
hello, i want this project to be able to combine voice to control multiple devices
Can we use esp home and add wakeup word and keep rest code same?
so did you solve the problem of slow response from Google Cloud Speech to Text ?
did you provide it with the optimal audio resolution ?
Hello bro the ChatGPT was paid can you give Gemini api key program for this same project and same connection
can you tell advantages to this project
How can I order a green board to assemble these kit.
Bro help me about Google api is there alternative credit card issues 😭
What changes do i need to make in the code to be able to use Google Gemini api instead of chatgpt api
How many words does it respond at a time?
Sir the api key of chatgpt is not free of cose. Please make a project using GEMINI Api key because it's completely FREE
Bro , did you find the answer??
@@DEBOJYOTI-X69 bro the video is made about 6 months ago. But I want a complete playlist on same topic
Unfortunately, it is using ir sensor to wake it up. You have to add wake up word in the project so that the assistant can wake up without ir sensor. _try to add it fast and provide us_
Easy 🙌🏻
@@saidevdhal hi can i get the code for text to speech and speech to text ,iam getting error from github code
@@Nk-creations_x Are you working on this project ? currently
What about proximity sensor ?
How it should be connected in circuit?
Also, how is working the way its changing color ?
same question , Have you found a solution to this issue?
You should check the last video in the V1 series. The connection diagram was there.
Project is amazing but I need help that speech to text works but text to speech doesn't work Plz give me suggestions. .??what should I do
sir where can i find ur pcb design so i can order it
Good work am replicating it but got issue with Speech_to_text. I followed video all steps but you didn't talk about network_param.h file. About "getting the certificate for your region". Both set of code uploadd fine for both esp32 but for the case is speechToText my esp32 crash (keeps on restarting when trying to communicate with Google STT api)
Error: Processing your Audio File
Json daata---
Guru Meditation Error: Core 1 panic'ed (LoadProhibited). Exception was unhandled.
Core 1 register dump:
PC : 0x4000c677 PS : 0x00060a30 A0 : 0x800d1d11 A1 : 0x3ffb1c90
A2 : 0x00000000 A3 : 0x3f4002cb A4 : 0x3ffb1d08 A5 : 0x000001d5
A6 : 0x00000000 A7 : 0x3ffb86e0 A8 : 0x3ff44000 A9 : 0x00000000
A10 : 0x00002a00 A11 : 0x3ff49040 A12 : 0x08000000 A13 : 0x00000007
A14 : 0xffffffff A15 : 0x0000000a SAR : 0x0000001e EXCCAUSE: 0x0000001c
EXCVADDR: 0x00000000 LBEG : 0x400014fd LEND : 0x4000150d LCOUNT : 0xffffffff
ELF file SHA256: 0000000000000000
Backtrace: 0x4000c677:0x3ffb1c90 0x400d1d0e:0x3ffb1cb0 0x400d2893:0x3ffb1f90 0x400d8054:0x3ffb1fb0 0x400899f6:0x3ffb1fd0
Rebooting...
Did you fix this error?
I am getting error into speech to text section actually I am not getting the text results what should I do
How to connect a display with it for see pictures
Do you need permission or patent right license for doing your gadget sales business to open public ?
How could I order the PCB board of the project
How much google cloud account charge initial amount when we verify.
Can we direct pay by netbanking that amount ?
The STT part runs fine but the TTS part gives me the error :
architecture or board not supported
check your esp version.
ESP Version - 2.0.14 for text to speech and 1.0.6 for speech to text
Great video, can you make a a video about esp32c6 and his zigbee ability inside arduino ide? There is no one video on that topic....
sir please upload fast and make a video on how to change the voice of ai and work like robot
G'day I am wondering if it can do Speech to Text, and how small can it really get? I saw the XIAO Esp32 being tiny.
I have a project that requires wireless, real-time Speech to Text capabilities, and 4 programmable buttons on it... in the smallest form factor possible... Max one index finger size.
Do you reckon it can be done with Esp32's or am I better off looking somewhere else?
what would you suggest?
what is the estimated budget
Why are you using two esp32's ??? It has dual core
Can u provide the PCB file so we can order from that website?
Actually I am a school student how do I create a pcb and get its so hard, any easy way to get pcb easily...?
Great video! When its finished and polished it will bean awesome unit to buy
Hi sir very useful this video
And then I am Studying college
I have submit my Minor project
In this Chatgpt based voice assistant project is ok or not
It was successful completed and ahow the output clearly pls clear my doubt and pls reply
Does anyone found anyway to increase recording time?
where did get pcb ?
Tumhe mila kya??
can you post pcb documents?
Thanks
What is the difference between alexa and this assistant bro give me clarity 😊
Can we make this project by capturing photos using esp32 cam according to our prompt and processing it using open ai and get the result through a speaker
No
Why didn’t you use esp32 module only
Sir can you provide Gerber file.. of the project.. please
where is pcb desion link
sir can you make evive - The Best DIY Prototyping Tool with in custam pcb desion and in limited buget it can fill all the dream who are intreasted in electroice plz help
Very good 😮
Thanks 😋
Can we replace google cloud ai with chtgpt api
How much the amount does google cloud take for verification ?
INR 2
@@techiesms Do they take money from my credit card after registration ,or do I have to pay money after the 2 INR?
@@nizaralshabibi3372that's automatic the will take it until the verification completes then they will refund
sir can you make a evive starter kit at low cost
Agar es assistance ko bna banaya kharide to kitne ka pad jayga
I wish i could solder, maybe if i can get the cornea transplant someday.
Please upload the 2.1 version as soon as possible 😅
WHICH SOFTWARE YOU USE TO MAKE PCB DESIGN AND ALSO CAN A BEGINNER USE THAT .....???
Easyeda
And yes it’s beginner friendly
In text to speech code i am getting an error in runtime which is memory.h no such file director is found plz help me out
how did you solve this ?
Download the directory
can you plz help me to this project plz give me your contact @@saidevdhal
What about rpi small version
can you make ESP32 project based on ollama Stand alone GPT
Hi your painless mesh library is not working pls check it😢
WiFiManager with SPIFFS data memory for the API keys
Can we train the chatgpt api to answer our questions in a particular way
Ofc
Isn't there free api instead of using Google cloud and OpenAI chatgpy
I can make one 😉
@@saidevdhal Did you come up with anyone?
@@Charliedave4332 yup
Why we need two ESP? We have multicore esp’s now, RTOS for multitasking, so why still use this two big guys?
They even not work at the same time, I also wonder why he used 2 ESP's..
Waiting for the Next Version video !!!!
can anyone tell me how i get the mother board
"Compilation error: 'i2s_read_bytes' was not declared in this scope " anyone help me please!!
Share the screenshot of the error on our WhatsApp
8200079034
Pretty sure this happens if you're using an esp32 board package version newer than 1.0.6. 'i2s_read_bytes' has been superseded in newer versions. If you downgrade to 1.0.6 using the board manager (see instructions in the video), that might help.
Exist code for esp32 voice recognition offline ?
can any on tell me how to get gerber file
PROCESSING YOUR AUDIO FILE
CONNECTION FAILED!
how do i fix this?
same here ! Did you find a solution ?
Did you have the PCB's Gerber file?
sir when i compile the code error: architecture or board not supported sd.h file ka sath
is there any solutionn??????
You found any solution for it ?
@@sarlasingh8851yes
Don't downgrade the devboard do the same with the latest version everything will run fine
first view ❤🎉
You're the best!
I love this channel.... But only thing which bother me the promotion of Chinese PCB manufacturers.
Your all videos are Chinese PCB manufacturers sponsored.... I know they are giant in this domain but try to promote Indian PCB manufacturers...Give preference to them. I can understand where it leads but nothing against your channel. we love the content. Just a humble request to promote Make in India or at least reduce Chinese content... Hope you flourish..cheers
Yes bro
I can understand your feelings.
But you don’t know the other side of content creation.
It is our job and it helps us to pay our bills. I’m not getting any kind of support from Indian PCB manufacturers or may be any Indian electronics company.
So if I stop Chinese collaborations as well, may be I need to shutdown my channel because I don’t have any other job. My RUclips channel is the only full time job I’m doing.
And only a RUclipsr will understand why I’m doing this collaborations, a normal people might not understand because even RUclips algorithm is very difficult to serve. I can observe that daily someone is shutting down their RUclips channel just because RUclips algorithm is not promoting their videos , which doesn’t generate views and doesn’t generate revenue as well.
That’s the only thing I can share with you and I hope you got some overview about what’s going behind.
In the end , I’ll just request you all people that whichever creators videos you genuinely like please try to like on it, comment on it and share it. That’s the best way to say thank you to creators like us.
And let’s hope in future I get some Indian Collaborations as well.
Hi techiesms ;) .. in case you still search for a ESP32 code to record audio with any length and store on SD card (2 sec .. 10 secs or minutes), let me know. I could finalize a code snippet myself for ESP32 not needing any Audio.h libraries). I also coded an SpeechToText API call (sending my wav from SD card to 'Deepgram' receiving an transcription text). Both togehter works perfect in my opinion. And will allow me to use one ESP32 only for the complete Assistant :) .. and funny: i might keep the 2 ESP still, because i love your Tx/Rx concept (my left ESP is doing Open AI only) .. my 2nd ESP is handling Recording & STT and TTS only (for any projects, not Open AI only)
Can you share me those codes on techiesms@gmail.com
I’m still looking for that solution
@@techiesms will do ;) .. give me few days to cleanup final code, Deepgram Dev folks might publish my STT piece in their Github too (will let you know when done, assuming in few days too).
@kaloprojects thanks, will be waiting for your response
Any update ?
@@techiesms yes, see my email ;)
Can I get the PCB design
yes pls
how to download garber file?
bro this is really great. but are you sure about this will it work 100%??????? why means i wants to buy hardware things
Try this
ruclips.net/video/zvR9DTfMwPE/видео.htmlsi=qXqF1ntvHHOkC0zq
bro i am can not order pcb
Mera nam kowsik, guntur ka bhAAi
Pcb?