> ## Documentation Index
> Fetch the complete documentation index at: https://docs.nomask.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# NoMask System Prompt

> Optionally prepend the same per-model persona used in the NoMask chat app.

> [!NOTE]
>
> Use a current public model ID from [Models](https://nomask.ai/models) or `GET /models`. Example IDs and options must match the selected model's advertised capabilities and limits.

By default the API is raw: your messages are the entire prompt and nothing is injected around them. One flag opts into the same persona the NoMask chat app uses.

## Raw by default

With no flags set, the model receives exactly the `messages` you send: no hidden system prompt, no identity, no behavioral instructions. What you see is what the model gets. This is the right mode when your application supplies its own system prompt.

## Enabling the NoMask persona

```json theme={"system"}
{
  "model": "nomask-1.2",
  "messages": [{"role": "user", "content": "Who are you?"}],
  "nomask_system_prompt": true
}
```

With `"nomask_system_prompt": true`, the per-model NoMask system prompt (identity, personality, and behavior rules for the selected model, as used in the chat app) is prepended as the first system message, ahead of your own messages. Your own system message still applies after it and can adjust tone and focus.

## Token cost

The persona adds roughly 800 tokens to the prompt, billed normally as input tokens. You will see the difference in `usage.prompt_tokens`. Leave the flag off if you do not need it.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.