Edge0

Overview

Edge0 is an open-source streaming MoE inference framework that runs large sparse AI models locally by streaming expert weights from SSD and predicting which experts will be needed next. It supports macOS iOS Android Linux and Windows with platform-dependent model and hardware support.

Key features

• SSD-streamed model weights
• Mixture-of-Experts inference
• Predictive expert routing
• Local AI inference
• OpenAI-compatible local server
• Cross-platform support
• Large model support