Enchanted Developers Only vs Pebble Mind

Enchanted is chat app for LLM researchers to chat with self hosted models. Enchanted supports Ollama API and all ecosystem models. It is necessary to have a running Ollama server to use this app and specify the server endpoint in app settings. Enchanted supports streaming and latest Chat API with conversation context. You can quickly switch between your models to test your prompts. Enchanted supports Markdown, therefore tables, lists and code will be nicely formatted. The app is in early days, but had a long roadmap to be leading tool for LLM community running self hosted models.

Your AI. Your models. Your device. Pebble Mind is a private, local-first AI workspace that lets you run powerful language and multimodal models directly on your device. Powered by llama.cpp, Pebble Mind brings local AI chat to iPhone and iPad without requiring an account, API key, or remote inference service. Download a compatible model, start a conversation, and keep your chats and AI processing on your device PRIVATE BY DEFAULT Your conversations, attachments, settings, and downloaded models are stored locally. AI inference with downloaded llama.cpp models happens on your device — your prompts do not need to be sent to a remote AI server. Web Search is optional and under your control. When enabled, Pebble Mind can access the web for current information and show Web Search and Sources activity inside your conversation. CHOOSE THE MODEL THAT WORKS FOR YOU Browse text and multimodal models from one model library. Pebble Mind shows download size, device compatibility, memory requirements, and availability to help you choose a model suited to your device. Download GGUF models for local use, switch between available models, and manage downloaded model storage from within the app. MORE THAN TEXT Compatible multimodal models can work with: • Photos • Camera captures • PDFs • Documents • Extracted text Use the same familiar chat experience whether you're asking a quick question, working with an image, analyzing a document, or continuing a longer conversation. BUILT FOR LOCAL AI • On-device inference powered by llama.cpp • Streaming AI responses • Text and multimodal model support • Searchable conversation history • Searchable model library • Optional web search with visible sources • Markdown-formatted responses • Stop and retry generation • Per-model Temperature, Top-P, Top-K, context length, and reply-length controls • Model download and storage management • Conversation sharing and export • Native, privacy-focused iOS experience PEBBLE MIND PRO The core Pebble Mind experience includes local AI chat and model browsing. Pebble Mind Plus is an optional one-time purchase — not a subscription — that unlocks additional features including custom model importing, expanded model management, advanced controls, generation metrics, data transfer tools, configurable search domains, and context compaction. NETWORK ACCESS, WHEN YOU CHOOSE IT Pebble Mind is designed to remain useful without remote AI inference. Internet access is used when you explicitly download models and when optional Web Search is enabled. Model files may be downloaded from Hugging Face, and web-search queries may be sent to external search services. Large AI models can require significant storage and memory. Pebble Mind shows model size and device suitability before download and provides a data/storage notice before the first large model download. IMPORTANT AI-generated responses can be inaccurate or incomplete. Pebble Mind should not be relied upon as a substitute for qualified professional advice or for medical, legal, financial, emergency, or other safety-critical decisions. Run AI locally. Choose your models. Keep control of your data.