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

# Otaku

> Using NanoGPT with Otaku, a roleplay terminal client

# Otaku Integration

[Otaku](https://otaku.sh) is a roleplay client for the terminal, available on macOS, Linux, and Windows through WSL. NanoGPT is built in, so your NanoGPT model catalog appears directly in the model picker.

## Setup

1. Install Otaku:

   ```bash theme={null}
   curl -LsSf https://otaku.sh/install.sh | sh
   ```

   Alternatively, install it with `uv tool install otaku`.

2. Run `otaku`. When the model picker opens, press **Tab** to move to the provider panel, select **NanoGPT**, and paste your NanoGPT API key into the `API key:` field. Otaku verifies the key and lists your NanoGPT model catalog on the left. Your API key is stored encrypted in the configuration file.

3. Choose a model and start chatting. The prompt shows `$` while a cloud model is selected.

## Useful commands

* `/model` (Ctrl+O) reopens the picker to switch models or providers.
* `/balance` shows your NanoGPT balance.
* `/context` shows exactly what is sent to the model.

## Manual configuration

To configure Otaku manually, add your API key to the `[nanogpt]` section of `~/.otaku/configs/providers.toml`:

```toml theme={null}
[nanogpt]
url = "https://nano-gpt.com/api/v1"
api_key = "your-nanogpt-api-key"
```

For more information, see the [Otaku guide and demo](https://otaku.sh) or the [Otaku source code](https://github.com/enclavum/otaku).