Glyph vs Local LLM runner

Glyph is a privacy-first local AI chat app for iPhone that helps you work with on-device models, personal context, and your own tools without turning every conversation into a cloud workflow. Chat with AI on your device, download local models, and organize your work with features designed for repeat use: - Private AI conversations with a clean chat experience - Local LLM support for offline-capable help on your iPhone - Knowledge bases for documents, notes, and reference material - Personas for custom assistant behavior and specialized workflows - Memories that help Glyph keep track of useful details over time - App-wide onboarding that explains what Glyph can do before you commit to an account Glyph is built around user control. You can continue without logging in, or sign in to sync account and subscription access with Fortaify. Free users can create up to 2 knowledge bases, 1 persona, and 5 memories. Glyph Pro unlocks expanded access to personas, knowledge bases, and memories across supported Glyph products. Use Glyph for local AI chat, research, drafting, brainstorming, coding help, document Q&A, personal organization, and privacy-focused AI workflows. When you use local models, your conversations and generated content are handled on your device by default. Terms of Use (EULA): https://www.apple.com/legal/internet-services/itunes/dev/stdeula/ Privacy Policy: https://www.fortaify.com/Privacy

Run LLM on your iPhone, 100% local, no network (Network needed once to download a model) Models: • Qwen2.5 0.5B Instruct — ~0.4 GB (4 GB RAM) • Llama 3.2 1B Instruct — ~0.8 GB (4 GB RAM) • Qwen2.5 1.5B Instruct — ~0.9 GB (4 GB RAM) • Llama 3.2 3B Instruct — ~1.9 GB (6 GB RAM) • Qwen3 4B Instruct 2507 — ~2.5 GB (6 GB RAM) • Qwen2.5 7B Instruct — ~4.4 GB (8 GB RAM) Also: • Add your own GGUF (for example, from Hugging Face) • Auto-estimated max context from GPU memory; you can change it Requires iOS 18 or later. Tested on iPhone 11, 14 Pro, 15 Pro with iOS 26. Qwen models © Alibaba Cloud. Licensed under Apache License 2.0. Llama 3.2 is licensed under the Llama 3.2 Community License, Copyright © Meta Platforms, Inc. All Rights Reserved. Built with Llama. On-device inference uses llama.cpp (MIT License).