# `TokenCount`

How many tokens a prompt or an answer holds, by its model's tokenizer: a whole number from 0.

[AI](types.md#ai); ordered; since Jadpo 0.1.0.

## Make one

Build one from a whole number from 0 to 9007199254740991, `TokenCount(0)`; a number outside them breaks `min` or `max`.

`text.parse(raw, as: TokenCount)` reads one from its text; text that isn't one at all fails `malformed`, such as `-1`, `1.5` or `1,200`.

## Text, JSON and storage

`text.from` writes its canonical text, the one form `text.parse` reads back. JSON sends it as a number.

| Text | JSON |
|---|---|
| `0` | `0` |
| `350` | `350` |
| `128000` | `128000` |

It is stored in one column named for the field, of type `bigint`. `350` is stored as `350`.

## Shown to people

`text.format(value, locale:)` shows it to people in their locale.

With `locale: Locale.en_gb`:

| Value | Shown |
|---|---|
| `0` | `0` |
| `350` | `350` |
| `128000` | `128,000` |

## Operations

It has the contract's own operations: `text.parse`, `text.from`, `text.format`, `json.read`, `json.write` and `==`. It is ordered, so `<` and sorting take it. [Every standard type meets one contract](../topics/every-standard-type-meets-one-contract.md) says how they work.