Open Source Impact
This repository provides an MLX-optimized, 4-bit quantized DeepSeek LLM for local inference on Apple silicon.
The model is publicly distributed via Hugging Face and used by members of the MLX community to run on-device LLM inference without relying on cloud APIs or CUDA-based stacks.
This project demonstrates practical hardware-aware ML optimization by leveraging Apple’s unified memory architecture and Metal GPU acceleration.
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support