Output truncated when calling the OpenAI API

I had some Python code written that called the OpenAI API. For some reason, during execution of this code I noticed that the output was truncated.

This was the snippet of the Python code that I had written that called the OpenAI API:

response = await openai.ChatCompletion.acreate (
        model=model,
        messages=messages,
        temperature=0.7,
        max_tokens=200,
        n=1,
        stop=None,
    )

The truncation of the output was due to the max_tokens parameter in my OpenAI API call. The max_tokens parameter limits the number of tokens (words and punctuation marks) in the response.

• The maximum allowed tokens for gpt-3.5-turbo is 4096.
• The maximum allowed tokens for gpt-4o is 124,000.

I was using gpt-4o and upped my max tokens to 4000 and my issue was resolved.

It is possible to look into the stream=True option if the response was too large, which was unnecessary in my case.