@@SimpleEarthSelfReliance Yes, that's the API of OpenAi. But for those who don't know much about it, it might be more intuitive to call it chatGPT. Again, you are right.
I'm waiting for the moment you use this library to send from an esp-cam an image to chatgpt asking information about it. It will be the general AI and you don't need very expensive hardware.
Totally agree with you! Certainly, I think it will be possible in the near future. It will enable more projects on low-performance MCUs. That's why I'm making and preparing this library.
Can the program's prompts be customized, or are they fixed? Additionally, is it possible to receive input from the serial port window? Please let me know. Thank you very much.
Like what you use at chat.openai.com/chat, you can ask what you want. After WiFi is connected, you can receive input from the serial port window and then toss it into the API.
Hi there! I wanted to make something similar to the ESP32 Berry for my school's final test, but I'm adding something more. Can you tell me where's the source code of this version with ChatGPT inclused? Please I don't have much time left and I need to buy the necessary parts. Thanks
@@ThatProject It won't be that much easy, especially for a 14 years old boy like me who's still discovering the programming world, but I'll try. Did you put some libraries or something in the SD card? Because the ESP32's flash memory isn't enough. I'm talking about the first case of the video
The exact project which i searching for the couple of day back and now i found it, thank you bro , can you try to create a real time chat bot with this chatgpt with speaker like alexa and siri..
Hi, can u please design a solution with esp32 i2s microphone, stt api , chatgpt api and i2s speaker? if we can talk with chatgpt it would be more fun...thx
ESP8226 on a mega2560, I read later in the GitHub issues that it could be a memory problem? With the full message I get x lines of json header before it starts giving a blank line. Do you think there is a way to write it immediately to the mega2560, would it overcome the problem?
@@fsowner This is definitely an issue due to insufficient memory. Because the String value of the result obtained from OpenAI's API is too long, it is cut off in the middle if memory is insufficient. Is it the same if you use the full_message function? The use of Arduino JSON parser requires more memory usage, so the full_message function may be good which is not using ArduinoJSON.
@@ThatProject Yes i am using the full_message function, you get a random number of lines of the header before it stops printing anything sadly. Do you think there is any solution? It should work on for example an ESP32?
Check More Projects - youtube.com/@ThatProject
Github Repository - github.com/0015/
Join FB Group - facebook.com/groups/138965931539175
I know I'm repeating myself, but I can not understand why you do not have more followers, you deserve it!!
Amazing job, i have made one lib by myself too!! excellent work
Very nice. Thanks
To be clear, ChatGPT is the frontend/GPT for OpenAI. What you're referring to is the OpenAI API, not ChatGPT.
@@SimpleEarthSelfReliance Yes, that's the API of OpenAi. But for those who don't know much about it, it might be more intuitive to call it chatGPT. Again, you are right.
Amazing project at the of the video.
I'm waiting for the moment you use this library to send from an esp-cam an image to chatgpt asking information about it. It will be the general AI and you don't need very expensive hardware.
Totally agree with you! Certainly, I think it will be possible in the near future. It will enable more projects on low-performance MCUs. That's why I'm making and preparing this library.
Can the program's prompts be customized, or are they fixed? Additionally, is it possible to receive input from the serial port window? Please let me know. Thank you very much.
Like what you use at chat.openai.com/chat, you can ask what you want. After WiFi is connected, you can receive input from the serial port window and then toss it into the API.
is there any example for arduino uno r4 wifi?
Working on it. github.com/0015/ChatGPT_Client_For_Arduino/issues/6
nice project..👍
Hi there! I wanted to make something similar to the ESP32 Berry for my school's final test, but I'm adding something more. Can you tell me where's the source code of this version with ChatGPT inclused? Please I don't have much time left and I need to buy the necessary parts. Thanks
Unfortunately, it's not finished yet. You have to implement it yourself with LVGL and ChatGPT Client library. It will be shared in another project.
@@ThatProject It won't be that much easy, especially for a 14 years old boy like me who's still discovering the programming world, but I'll try. Did you put some libraries or something in the SD card? Because the ESP32's flash memory isn't enough. I'm talking about the first case of the video
@@WalidsTech I made it to read and write text files from SD card. You don't have to worry about it at all.
The exact project which i searching for the couple of day back and now i found it, thank you bro , can you try to create a real time chat bot with this chatgpt with speaker like alexa and siri..
I think it's possible if we add some features one by one. haha
wow!
Hi, can u please design a solution with esp32 i2s microphone, stt api , chatgpt api and i2s speaker? if we can talk with chatgpt it would be more fun...thx
this sample looks like not works with nodemcu esp12
I updated this library to support ESP8266. Check it out with the new version, 0.1.1 (It takes time to get it via Library Manager though).
@@ThatProject it's works~thank you
Is this still working, it's not working for me..
Which MCU are you using?
ESP8226 on a mega2560, I read later in the GitHub issues that it could be a memory problem? With the full message I get x lines of json header before it starts giving a blank line. Do you think there is a way to write it immediately to the mega2560, would it overcome the problem?
@@fsowner This is definitely an issue due to insufficient memory. Because the String value of the result obtained from OpenAI's API is too long, it is cut off in the middle if memory is insufficient. Is it the same if you use the full_message function? The use of Arduino JSON parser requires more memory usage, so the full_message function may be good which is not using ArduinoJSON.
@@ThatProject Yes i am using the full_message function, you get a random number of lines of the header before it stops printing anything sadly. Do you think there is any solution? It should work on for example an ESP32?
@@fsowner It works well on ESP32. There is no special solution as of now. sorry.