Advanced Settings

Stream Data

Default - ON. When ON, output from the LLM is shown word-by-word, as it is produced. If you'd prefer the output to appear all at once when the LLM is done thinking, turn this feature OFF.

Safe Context Window

Default - OFF. When ON, context will be automatically reduced to the model's maximum context size. If OFF, you will get an error message if your context gets too big. Be careful--turning this on means you might unexpectedly lose meaningful context without a warning!

Charts

To maintain consistency, chart generation is now handled exclusively by the dedicated StackAI analysis tool.

Data & Time

Default - ON. When ON, the LLM will be aware of the current date & time.

Guardrails

Default - OFF. Turn ON guardrails to screen for toxic content, legal advice, or suicidal thoughts.

PII Compliance

Default - OFF. Turn ON PII Compliance in order to hide certain types of PII from the LLM if entered by a user.

Temperature

Default - 0. Increase temperature in order to increase randomness in the output, making the model less deterministic.

Max Output Length

Default - 3000. Increasing this slider will allow the model to give more verbose answers.

Retry on Failure

Default - OFF. Turn ON to enable retrying when execution of a node fails. This can help make your project more robust

LLM Fallback Mode

Default - OFF. Turn ON to make your project more robust to model failure. This allows you to select a backup model and provider to use in case something goes wrong.

Fallback Branch

Default - OFF. When turned ON, you can specify a different flow to follow in case this LLM node's execution fails.