feat(local-apps): Add Nexa SDK integration

Open mengshengwu opened this issue 2 months ago • 1 comments

Introduction

Hello! Friends from Huggingface 👋

Thank you for maintaining this amazing library and ecosystem.
Following a recent discussion with Hugging Face (see here), your CEO expressed interest in exploring collaboration to bring on-device model inference to the community. This PR proposes adding Nexa SDK to the local apps section as a new runtime integration option.

About Nexa SDK

Nexa SDK is an on-device inference framework that runs any model on any device, across any backend. It runs on CPUs, GPUs, NPUs with backend support for CUDA, Metal, Vulkan, and Qualcomm / Intel / AMD NPU. It handles multiple input modalities including text 📝, image 🖼️, and audio 🎧. The SDK includes an OpenAI-compatible API server with support for JSON schema-based function calling and streaming. It supports model formats such as GGUF, MLX, Nexa AI's own .nexa format, enabling efficient quantized inference across diverse platforms.

Example Demos

🖼️ Multi-Image Reasoning Spot the difference across two images in multi-round dialogue.	🎤 Image + Text → Function Call Snap a poster, add a voice note, and the AI agent creates a calendar event locally.
🎶 Multi-Audio Comparison Identify differences between two music clips — fully offline.