Run small GGUF chat models fully offline on iOS and Android.
Phone RAM limits model size more than the app does.
PocketPal AI runs quantized GGUF models directly on iPhone and Android with no server and no account. It is the honest answer to “ChatGPT without the cloud” on a phone. Stick to 1B–3B models.
on-device GGUF · iOS and Android · model import · no server