What settings can I change in the Playground?

The settings panel in the Playground lets you shape how the model responds:

• Model: choose which model to use.
• System instructions: set the model's role or rules for the conversation.
• Temperature, Max tokens, and Top P: tune the response. Higher temperature is more varied; max tokens caps response length; Top P adjusts the diversity of word choices.
• Reasoning effort: choose how much a reasoning-capable model reasons before answering (see Adjusting reasoning).
• Streaming: stream the response as it's generated, or wait for the full reply.
• Advanced: add Additional messages to seed the conversation, define a JSON schema for structured output, or add Tools.

Select Reset parameters to return the settings to their defaults.

Two tabs help you go further: Compare runs the same prompt against more than one model side by side, and Get code turns your current setup into ready-to-run code.

Usage and limits: requests you make in the Playground count toward your team's usage and rate limits, the same as API calls.