help me with gpt-5 call

[jsmith173]

I'm using LangChain to make OpenAI API calls, and it works fine with the GPT-4o model.

Here’s the relevant part of the code:

client = OpenAI(
    base_url=base_url,
    api_key=api_key,
)

response = client.chat.completions.create(
    model=self.config['model_actual'],
    messages=self.messages,
    stream=stream,
    n=1,
    frequency_penalty=0,
    presence_penalty=0,
    seed=12345,
    top_p=0.1,
    max_tokens=max_tokens
)

If the model is set to GPT-5, the following error message appears:
Error message:

"Unsupported parameter: 'max_tokens' is not supported with this model. Use 'max_completion_tokens' instead."

Please help me out.

[NatashaKSS]

You're probably almost there! Like what the error says, try changing max_tokens to max_completion_tokens instead. max_tokens is a variable you've defined somewhere in your Python code and you just gotta pass it to the right parameter.

client = OpenAI(
    base_url=base_url,
    api_key=api_key,
)

response = client.chat.completions.create(
    model=self.config['model_actual'],
    messages=self.messages,
    stream=stream,
    n=1,
    frequency_penalty=0,
    presence_penalty=0,
    seed=12345,
    top_p=0.1,
    max_completion_tokens=max_tokens
)

[NatashaKSS]

Can't tell what model you're using but maybe this discussion https://community.openai.com/t/why-was-max-tokens-changed-to-max-completion-tokens/938077 will be helpful for you in understanding why that parameter got changed

[RaschidJFR]

I'm getting the same error on GPT-5 in JS despite using maxCompletionTokens.

const gpt5 = new ChatOpenAI({
  modelName: 'gpt-5',
  maxCompletionTokens: 1000,
  apiKey: OPENAI_API_KEY,
  temperature: 0.1,
});

[Error: 400 Unsupported parameter: 'max_tokens' is not supported with this model. Use 'max_completion_tokens' instead.]

This problem does not occur for gpt-4o-mini.

[HenningThorsvoll1]

Did you solve it