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

# What's Nito

> Any leading model, incognito, in the tool you already work in.

## More Models

Open-weight models like GLM 5.3, Kimi K3, DeepSeek V4 Flash, and Qwen3.8, and frontier models like Claude Opus 5, GPT-5.6, Gemini 3.7 Flash, and Grok 4.6, without changing your subscription, your setup, or where you work. New models are added as they release. See [Models](/models).

## Incognito

Models can be selected at different privacy levels, which carry different, cumulative incognito protections.

| Level | Description (each level builds on the last) |
| :- | :- |
| **Anonymous** | Frontier models, with no identity attached to the request. |
| **Private** | Open-weight models with zero data retention upstream. |
| **Confidential** | Runs inside sealed hardware, with attestation you can verify on every request. |
| **Encrypted** | Runs inside the same sealed hardware as Confidential, with prompt and response encrypted on your device on the way in and out. In the Nito Browser App and the Nito API today, coming soon to the Nito Plugin. |

Whatever the level, **Nito stores nothing you send or receive.** What each level protects: [How Privacy Works](/privacy/overview).

## Three Ways In

* **[Nito Browser App](/web-app).** Chat with any leading model at app.trynito.ai. Switch models per chat, run one prompt across several models with Model Fusion, attach files, and keep your history encrypted in your browser or in an incognito chat that is never saved.
* **[Nito Plugin](/plugin).** Inside Claude Code or Codex. Your harness keeps its own model; Nito adds commands to ask another model, send a prompt your harness never sees, or run a whole session through Nito.
* **[Nito API](/api).** Endpoints compatible with the OpenAI and Anthropic SDKs, plus MCP. One key reaches every leading model, incognito.

One account, one plan, usage shared across all three.

## Receipts

Every reply carries a record of what ran: the model, the privacy level, and the token counts. In the Browser App, Confidential and Encrypted replies also carry the attestation evidence. The receipt never contains your prompt or the model's response. See [Receipts and Proof](/web-app/receipts-and-proof).

## Getting Started

<CardGroup cols={3}>
  <Card title="Browser App quickstart" icon="window" href="/web-app/quickstart">
    Get started in minutes.
  </Card>

  <Card title="Install the plugin" icon="terminal" href="/plugin/get-started/what-the-plugin-is">
    Claude Code and Codex.
  </Card>

  <Card title="API quickstart" icon="code" href="/api/get-started/quickstart-60-seconds">
    One key, one request.
  </Card>
</CardGroup>


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