---
slug: "pi-gpt-fast"
source_type: "readme"
source_url: "https://cdn.jsdelivr.net/gh/10ego/pi-gpt-fast@main/README.md"
repo: "https://github.com/10ego/pi-gpt-fast"
source_file: "README.md"
branch: "main"
---
# pi-gpt-fast

Tiny pi package that adds a `/gpt-fast` toggle for OpenAI GPT Fast mode.

When enabled, eligible OpenAI/OpenAI Codex GPT requests are sent with:

```json
{ "service_tier": "priority" }
```

## Install

```bash
pi install npm:pi-gpt-fast
```

Or try it for one run:

```bash
pi -e npm:pi-gpt-fast
```

For local development, replace the package name with a checkout path such as `./pi-gpt-fast`.

## Usage

```text
/gpt-fast          # toggle on/off
/gpt-fast on       # enable fast mode
/gpt-fast off      # disable fast mode
/gpt-fast status   # show current state
```

You can also start pi with Fast mode enabled:

```bash
pi --gpt-fast
```

Fast mode only applies to OpenAI/OpenAI Codex GPT models. Unsupported models/accounts may ignore or reject the priority service tier.
