You can access the GPT backends with straight forward curl commands.

    curl https://api.openai.com/v1/completions \
      -H "Content-Type: application/json" \
      -H "Authorization: Bearer $OPENAI_API_KEY" \
      -d '{
      "model": "code-davinci-002",
      "prompt": "# Python 3 \ndef remove_common_prefix(x, prefix, ws_prefix): \n    x[\"completion\"] = x[\"completion\"].str[len(prefix) :] \n    if ws_prefix: \n        # keep the single whitespace as prefix \n        x[\"completion\"] = \" \" + x[\"completion\"] \nreturn x \n\n# Explanation of what the code does\n\n#",
      "temperature": 0,
      "max_tokens": 64,
      "top_p": 1,
      "frequency_penalty": 0,
      "presence_penalty": 0
    }'

Access to the systems doesn't need to be done through the ChatGPT front end - that's a neat technology demo.