Favicon of Personal LLM

Personal LLM

A local AI chat app for iOS and Android runs GGUF models and answers photo and document questions offline after downloads, without an account.

Screenshot of Personal LLM website

Personal LLM runs AI chat on your phone for people who want to ask questions without sending them to a cloud model. Chat, photo analysis and document questions work offline after you download a model and any required image support files.

  • Model choice: Choose Qwen, Gemma, GLM or Ministral from the catalog, or add a GGUF model by URL with an optional vision projector for image input.
  • Photos and documents: Ask about images or attach PDF, text and Markdown files; the app searches document text on your phone and identifies the passages used in its answer.
  • Chat controls: Switch models within a conversation, set a system prompt for each chat and adjust response settings; supported reasoning models can display their working in a collapsible panel.
  • Read aloud and backups: On-device text-to-speech reads answers aloud, and file exports let you back up and restore chats and settings.

The app runs on iOS and Android through llama.cpp. It uses Metal on iPhone and OpenCL on supported Snapdragon Adreno 700+ phones, with a CPU fallback. Model cards assess compatibility using your phone's RAM. The smallest catalog model needs at least 2 GB; other catalog models require 4, 6, 8 or 12 GB, depending on the model.

It's free to use with ads and also has paid options. Inference runs locally, but model downloads from Hugging Face, Google AdMob ads and purchase checks use the internet. AdMob may collect device information. Chats and attached photos stay on the phone; reporting an answer to support sends that answer and your note. The app has no cloud sync and excludes its files from iCloud and Google backups.

Similar to Personal LLM