Mano-P Local Model & Cider SDK Are Live — Run GUI Automation Fully On-Device

[fanyue0203-glitch]

Hey everyone 👋

We've released the Mano-P 1.0-4B local model and the Cider inference SDK — you can now run Mano-P's full GUI automation capability on your own Mac, with zero data leaving your device. 🔒

What's new

🤖 Mano-P 1.0-4B Local Model
Runs on Mac mini / MacBook with Apple M4+ & 32GB+ RAM

Download:
🤗 HuggingFace
🪄 ModelScope

⚡ Cider — INT8 Activation Quantization SDK for MLX
Fills the gap MLX doesn't cover: W8A8 and W4A8 online activation quantization
1.4x–2.2x prefill speedup vs MLX native W4A16 on Apple M5 Pro
~80 tokens/s decode with Mano-P 4B
Works with any MLX model, not just Mano-P 🎯

Source: github.com/Mininglamp-AI/cider

🏗️ Mano-AFK — Build Apps From a Single Sentence
End-to-end automated app construction: natural language → requirements → code → deploy → E2E GUI test → fix → ship ✅
E2E testing powered by local Mano-P — screenshots never leave your machine

⚠️ On 16GB devices, W8A8 doubles weight memory — recommend 4GB+ free beyond model size for full benefit

Source: github.com/Mininglamp-AI/mano-afk

Feedback and issues welcome. Let us know what you build with it! 💬🚀