1 min read
Adjust temperature and token limit
Set the generation options offered by the agent.
Before you start
- Use the panel as a platform administrator; the AI Agents list redirects other profiles.
- Have an existing agent.
Step by step
Open Configuration and adjust Temperature from zero to one.
Enter Max Tokens within the displayed range of 256 to 4096.
Save and compare responses in Test using representative questions.
Check your result
The saved options are sent to the model.
What you need to know
- Temperature does not guarantee factual accuracy. The limit is in tokens, not words.
- The provider may apply its own rules for accepting these options.