AILY — 多引擎 AI 助手 / AILY AI Assistant

AILY runs state-of-the-art large language models directly on your iPhone — no internet connection required. In airplane mode, on the subway, or during overseas travel without a cellular plan, your personal AI keeps answering with zero latency and zero per-message cost, ensuring your thoughts never leave your device.

Powered by llama.cpp on-device inference, AILY features a curated 20-model offline library spanning from 0.5 GB pocket models to 4 GB+ flagships (Qwen, Gemma, Llama, Phi, SmolLM, DeepSeek distills), complemented by intelligent region-routing downloads, custom GGUF file imports, and an active memory guard that prevents system thrashing. It also integrates Apple Intelligence on-device models for instantaneous, native reasoning.

When frontier-level reasoning is required, simply bring your own API keys to connect Claude, GPT, DeepSeek, Kimi, GLM, Qwen, Gemini, Grok, or a local Ollama server. Keys are secured in the iOS Keychain and talk directly to official provider endpoints with no intermediary proxies. With a pay-once lifetime architecture, vision photo input, and absolute zero tracking, AILY makes AI truly personal and sovereign.