Response from the model for a grounded answer.


Candidate answer from the model.

answerable_probability float

Output only. The model's estimate of the probability that its answer is correct and grounded in the input passages.

A low answerable_probability indicates that the answer might not be grounded in the sources.

When answerable_probability is low, some clients may wish to:

  • Display a message to the effect of "We couldn’t answer that question" to the user.

  • Fall back to a general-purpose LLM that answers the question from world knowledge. The threshold and nature of such fallbacks will depend on individual clients’ use cases. 0.5 is a good starting threshold.