Top-P

What it means

A sampling setting that limits next-token choices to likely candidates whose combined probability reaches a specified threshold.

Code example

const res = await openai.chat.completions.create({
  model: "gpt-4o",
  top_p: 0.9, // only sample the most likely tokens
  messages: [{ role: "user", content: "Write a tagline" }],
});

Why it matters

It provides another way to control response generation and output diversity.