Questions are keyed in a map, so each answer comes back under the name you gave it. Choice selects one option from a defined set and returns the winning option, the full probability distribution and a confidence value, with up to 255 options allowed. Score evaluates content against ordered, descriptive levels, minimum 2 and maximum 10, and returns a probability-weighted mean of the level numbers. Noul returns a single number between 0 and 1, the probability that a yes/no statement is true, and no separate confidence value.
The docs advise asking “the most explicit, narrow, specific, atomic questions you can” and breaking complex questions into separate ones that each evaluate a single property. That is cheap to do, because all questions in a call are evaluated in parallel and adding more typically does not add latency.
A worked example from the docs: a bug-severity Score with the levels “Cosmetic; no impact to functionality”, “Broken or degraded feature, but workaround exists” and “Blocking issue; no workaround exists” returned a score of 1.30 at confidence 0.54, from probabilities of 0.0, 0.70 and 0.30.
TypeSafe does not document a maximum number of questions per request, so do not assume one. More on the answer shapes in Choice, Score and Noul, and on what questions run against in state.