Max tokens vs max_completion_tokens - which to use?

Jordan LeeMar 15, 2026

I'm confused about the difference between max_tokens and max_completion_tokens in the API. The docs mention both but I'm not sure when to use which.

# Which one should I use?
response = client.chat.completions.create(
    model="gpt-4o",
    messages=[...],
    max_tokens=1000,           # This one?
    # max_completion_tokens=1000,  # Or this one?
)

Are they the same thing? Is one deprecated?

2.3k views8 replies19 likesSolved

1 Reply

Jessica LiuStaffAccepted AnswerSep 17

The validation error is likely because the tools field needs to be at the same level as messages, not nested inside. Here's the correct format:

{
  "messages": [...],
  "tools": [{"type": "function", "function": {"name": "get_weather", ...}}]
}

Make sure each training example includes the complete tool definition.