Cameron Armstrong’s AmigaGPT just hit version 2.4.0, and for those running AmigaOS 3.1 or higher — including 4.1 and MorphOS — it’s a welcome update. This release packs a mix of practical upgrades and behind-the-scenes improvements that make using OpenAI’s models on an Amiga smoother and a bit more fun.
The most notable change is the addition of an ARexx port, which opens the door for controlling AmigaGPT via scripts or other applications. For power users and tinkerers, this means less clicking around and more automation — a throwback to how Amiga users have always gotten things done efficiently.
There’s also a fix for the About window — a small detail, but one that matters if you actually want to read all the information displayed.
Support has been added for OpenAI’s latest chat and text-to-speech models, including the newest voices and control options. Users of the gpt-4o-mini-tts model can now modify how the voice sounds through direct instructions — think changing tone, speed, or inflection with simple prompts.
AmigaGPT’s chat interface now uses the updated Responses endpoint instead of the older Chat Completions system. This change isn’t flashy on the surface, but it gives the tool a more future-friendly backend, setting it up for upcoming improvements and features without needing another overhaul.
For those unfamiliar, AmigaGPT is a ChatGPT client that fits right into the Amiga environment. Whether you’re using it for writing, coding help, creative ideas, or just seeing what AI can do, it makes use of OpenAI’s text and image generation models — yes, even DALL-E — while keeping the classic Amiga vibe intact. It’s fast, responsive, and works with speech output too.
If you want a closer look at how the update works in practice, or just want to see AmigaGPT in action, check out the accompanying video. It gives a better sense of how the new features feel on a real Amiga setup.
