Gadeon vs FrameReply

Gadeon runs Qwen3.5 and QwenPaw hybrid language models entirely on your device's Apple Neural Engine, using Core ML as its only framework dependency. There's no C, no Python, and no cloud calls: every token is generated locally, keeping your conversations private. Under the hood, these models mix Gated DeltaNet recurrent layers with a smaller number of attention layers. Gadeon compiles that hybrid architecture directly onto the Neural Engine's dataflow accelerator, so both prompt processing (prefill) and token generation (decode) run on-chip instead of falling back to the CPU or GPU. On an Apple Silicon Mac, prefill measures roughly 2 to 2.4 times faster and about 4.4 times more energy-efficient per token than a Metal GPU baseline, while leaving the GPU free for other work. Conversations in Gadeon are append-only: each turn extends the model's fixed-size recurrent state instead of re-processing the entire chat history, so response latency doesn't grow as a conversation gets longer. The app can also roll back to an earlier point in a conversation, dropping reasoning traces or replaying tool calls in a clean form without reprocessing everything from scratch. For ternary-quantized models, Gadeon switches to a pure-Swift Metal GPU backend behind the same session and tokenizer interface, so both compute paths feel identical from the chat UI. Gadeon is a proof of concept for bringing Gated DeltaNet hybrid models to the Neural Engine through Core ML - built to explore what's possible on Apple's own silicon.

FrameReply is an open-source chat assistant that delivers two context-aware reply ideas directly on your current messaging screen using Shortcuts and snippets. It remembers useful details about each person, supports custom personas, and adapts to your writing style. STAY IN THE CONVERSATION • Get context-aware reply ideas in an interactive Shortcuts overlay without leaving your current messaging screen. • Bring in chat screenshots or copied messages via Share Sheet or Shortcuts. The image shortcut can also capture the current screen or launch with Back Tap. KEEP THE THREAD • Automatically add new messages to an existing chat when there is a reliable match, or add them directly to one you already have open. • Memorize useful person-specific context, such as preferences, shared history, and recurring plans. • Set a Current Goal for the chat, or add one-time Reply Guidance for what you want to say next. MAKE IT SOUND LIKE YOU • Choose a built-in persona or create your own. Personas can learn recurring patterns from messages you send. • Get Conversation Strategy and Strategy Rationale that explain why the replies were suggested. YOU STAY IN CONTROL • AI suggestions are drafts. The app never sends messages for you; you choose what to copy and whether to edit or send it in your messaging app. • Delete a chat, remove a provider connection, or erase all local data from your device whenever you choose. REQUIREMENTS • iPhone with iOS 26 or later. • Your own API key for OpenAI, OpenRouter, or MiniMax. • Provider usage may incur charges from that provider.