How to call the ChatGTP API from R (in 2023-03-01)

chat-gtp-api-call.R

# How to call the new (as of 2023-03-01) ChatGTP API from R
# Get your API key over here: https://platform.openai.com/
api_key <- "sk-5-your-actual-api-key-Fvau6" # Don't share this! 😅

library(httr)
library(stringr)

# Calls the ChatGTP API with the given promps and returns the answer
ask_chatgtp <- function(prompt) {
  response <- POST(
    url = "https://api.openai.com/v1/chat/completions", 
    add_headers(Authorization = paste("Bearer", api_key)),
    content_type_json(),
    encode = "json",
    body = list(
      model = "gpt-3.5-turbo",
      messages = list(list(
        role = "user", 
        content = prompt
      ))
    )
  )
  str_trim(content(response)$choices[[1]]$message$content)
}


# For example: 
ask_chatgtp("What function makes a histogram in R?")
## "The `hist()` function makes a histogram in R."

I've got a couple of suggestions:

1. Return (return()) the actual OpenAI's response so that if we want to store the returned results or check at anything else, we can simply assign it to an object and use it later on.
2. Add a cat() to the print so it renders nicely.
3. Warn with the returned error message if that's the case

This should fix all:

  ret <- content(response)
  if ("error" %in% names(ret)) warning(ret$error$message)  
  if ("message" %in% names(ret$choices[[1]]))
    cat(stringr::str_trim(ret$choices[[1]]$message$content))
  return(invisible(ret))

@laresbernardo I agree rendering cud actually help greatly, right now it just gives everything in a messy order. where do i put ur fix in the whole code ?

# How to call the new (as of 2023-03-01) ChatGTP API from R
# Get your API key over here: https://platform.openai.com/
api_key <- "sk-5-your-actual-api-key-Fvau6" # Don't share this! 😅

library(httr)
library(stringr)

# Calls the ChatGTP API with the given promps and returns the answer
ask_chatgtp <- function(prompt) {
  response <- POST(
    url = "https://api.openai.com/v1/chat/completions", 
    add_headers(Authorization = paste("Bearer", api_key)),
    content_type_json(),
    encode = "json",
    body = list(
      model = "gpt-3.5-turbo",
      messages = list(list(
        role = "user", 
        content = prompt
      ))
    )
  )
  ret <- content(response)
  if ("error" %in% names(ret)) warning(ret$error$message)
  if ("message" %in% names(ret$choices[[1]]))
    cat(stringr::str_trim(ret$choices[[1]]$message$content))
  return(invisible(ret))
}

# For example: 
ask_chatgtp("What function makes a histogram in R?")
## "The `hist()` function makes a histogram in R."

Thank you... This works now.

Starting from today content(response) returns an Error in UseMethod("content", x) : no applicable method for 'content' applied to an object of class "response".
Any ideas how to fix it?

Thank you!

@staceymir Stacey it is working for me.
Try restarting.

In the meantime look at this option
https://github.com/jcrodriguez1989/chatgpt

Thank you for the link pvsundar, I'll give it a try.

In case if anyone can help with the issue, here is a code:

library(httr)
api_key <- "..."
response <- POST(
url = "https://api.openai.com/v1/chat/completions",
add_headers(Authorization = paste("Bearer", api_key)),
content_type_json(),
encode = "json",
body = list(
model = "gpt-3.5-turbo",
messages = list(
list(
role = "user",
content = "What is ChatGPT?"
)
)
)
)
content(response)

I updated the httr package and now the code is working :-)

Thank you. By the way, when looking for internet essay samples to assist me write my essay, I came across this website https://www.topessaywriting.org/samples/culture On this website, I may get numerous free essay samples to assist me enhance and personalise my essay.