# `TypeSafeAPI.Answer.Noul`
[🔗](https://github.com/typesend/typesafe_ai/blob/v0.1.0-alpha.3/lib/typesafe_api/answer/noul.ex#L1)

A decoded Noul answer: the model's probability that the answer is yes.

`noul` is between 0.0 and 1.0 — a response outside that range is rejected as
an `:unexpected` error rather than clamped, so a malformed value cannot
quietly reach `TypeSafeAPI.Answer.gate/2`. Values near the extremes mean the
model is confident; values near 0.5 mean it isn't.

`confidence` is `max(noul, 1 - noul)`. It is this library's convention, not
something the model reported: the API sends no confidence for a Noul answer,
only the probability. It reads distance from 0.5 as certainty, so 0.92 and
0.08 both give 0.92, and it never drops below 0.5. It is not the same
measurement as a Choice or Score `confidence`, so do not tune one threshold
against numbers from the other. The field is here so the three answer structs
have the same shape; `TypeSafeAPI.Answer.confidence/1` returns the same value.

Because the value has a floor of 0.5, `TypeSafeAPI.Answer.gate/2` cannot
return `:escalate` for a Noul answer unless `:review` is above 0.5 — it
raises on a lower threshold rather than offering an unreachable band. Use
`TypeSafeAPI.Answer.yes?/2` for the yes/no reading and `gate/2` for routing.

# `t`

```elixir
@type t() :: %TypeSafeAPI.Answer.Noul{
  confidence: float(),
  id: TypeSafeAPI.Keys.key(),
  noul: float()
}
```

---

*Consult [api-reference.md](api-reference.md) for complete listing*