Nexa Android SDK Demo App

📣 NexaSDK for Android is highlighted by Qualcomm blog as "a simple way to bring on-device AI to smartphones with Snapdragon"

Overview

The Nexa AI Android SDK enables on-device AI inference for Android applications with NPU acceleration. Run Large Language Models (LLMs), Vision-Language Models (VLMs), Embeddings, Speech Recognition (ASR), Reranking, and Computer Vision models on Android devices with support for NPU, GPU, and CPU inference.

📖 For full documentation, see Android SDK Doc.

Device Compatibility

Supported Hardware

Minimum Requirements

• Android API Level 27+ (Android 8.1 Oreo)
• Architecture: ARM64-v8a

Quick Start with APK

Standard Demo App

Download and install the pre-built APK:

# Download: https://nexa-model-hub-bucket.s3.us-west-1.amazonaws.com/public/android-demo-release/nexaai-demo-app.apk
adb install nexaai-demo-app.apk

GPT-OSS NPU Demo

For running GPT-OSS model on Qualcomm NPU:

# Download: https://nexa-model-hub-bucket.s3.us-west-1.amazonaws.com/public/nexa_sdk/huggingface-models/gpt-oss-android-demo/nexaai-gpt-oss-npu.apk
adb install nexaai-gpt-oss-npu.apk

Build from Source

🎬 Watch the tutorial video showing how to run the sample app in 40 seconds.

Step-by-Step Instructions

1. Clone the repository

   bashCopy

   git clone https://github.com/NexaAI/nexa-sdk/
   

2. Open in Android Studio

   Open the bindings/android folder in Android Studio.

3. Download a model

   Follow the Android SDK Doc to download a model. Below are some examples to download:

   • Granite-4.0-h-350M-NPU
   • Granite-4-Micro-NPU
   • OmniNeural-4B

   Place the model in the app's data directory:

   Copy

   /data/data/com.nexa.demo/files/models/<model-name>
   

4. Build and run the app in Android Studio

Tutorial: LFM2-24B-A2B-Preview-GGUF

This walkthrough uses the LFM2-24B-A2B-Preview-GGUF model in the demo app.

1. Install the app
   Install the demo app (APK or build from source as above).

2. Select the model
   Open the model selector (dropdown next to the model name) and choose LFM2-24B-A2B-Preview-GGUF.

3. Download
   Tap Download to fetch the model to your device. Wait until the download finishes.

4. Load
   Tap Load. A load model config dialog appears: choose CPU, GPU, or NPU (for Qualcomm NPU), then tap SURE. Once the model is loaded, the chat area becomes available.

5. Chat
   Type your message in the input field at the bottom, then tap Send to get a response. Use Clear to clear the input or conversation as needed.