Snapdragon

Top Builders

Explore the top contributors showcasing the highest number of app submissions within our community.

Snapdragon

Snapdragon is Qualcomm's brand for its system-on-chip (SoC) family, combining processing, graphics, neural processing, modem, and radio connectivity on a single piece of silicon. From the Snapdragon 8 Elite powering flagship Android smartphones to the Snapdragon X Elite running Windows Copilot+ PCs, the platform targets the full spectrum of connected devices. Snapdragon is the primary hardware platform for on-device AI in consumer electronics, with the Hexagon NPU achieving up to 100 TOPS on the latest generation.

General
Developer	Qualcomm
Type	System-on-Chip Platform
License	Proprietary (silicon); developer tools include open-source components
GitHub	qualcomm/ai-hub-models
Documentation	Qualcomm AI Hub Docs
AI Hub	aihub.qualcomm.com

Core Features

Hexagon NPU — dedicated neural processing unit optimized for AI inference; achieves 40 TOPS (8 Gen 3) to 100 TOPS (8 Elite); ranked #1 on MLPerf Inference v4.0 mobile benchmark.
Heterogeneous AI Engine — workloads dynamically distributed across Hexagon NPU, Adreno GPU, and Oryon/Kryo CPU for optimal power efficiency.
Qualcomm AI Hub — cloud platform with 175+ pre-optimized models and remote profiling on 50+ real Snapdragon devices; install with pip install qai_hub_models.
Multi-precision support — FP32, FP16, INT16, and INT8 inference; enables quantized LLM deployment on mobile hardware.
On-device LLM execution — validated with Llama, Mistral, Qwen3, IBM Granite, and Stable Diffusion variants via AI Hub.
Sensing Hub — ultra-low-power always-on processor for continuous sensor monitoring without waking the main cores.

Snapdragon Platform Variants

Variant	Segment	NPU Performance	Notable Devices
Snapdragon 8 Elite	Flagship mobile (2024-2025)	100 TOPS	Samsung Galaxy S25 series
Snapdragon 8 Gen 3	Flagship mobile (2023-2024)	~40 TOPS	Samsung Galaxy S24 series
Snapdragon X Elite	Windows laptops / Copilot+ PC	45 TOPS	Microsoft Surface Laptop
Snapdragon X2 Elite	Windows laptops (2025)	80 TOPS	Next-gen Copilot+ PCs
Snapdragon AR2	AR glasses	Ultra-low power	Various XR wearables

Tools and Resources

Qualcomm AI Hub — aihub.qualcomm.com — browse pre-optimized models, run cloud profiling sessions, and download deployment artifacts.
AI Hub Models (GitHub) — github.com/qualcomm/ai-hub-models — open-source Python library (BSD-3-Clause) with 200+ model recipes.
AI Engine Direct SDK (QNN) — low-level SDK for building applications directly targeting the Hexagon NPU.
AI Model Efficiency Toolkit (AIMET) — quantization, pruning, and compression tools for edge-friendly model preparation.
Snapdragon Neural Processing Engine (SNPE) — runtime that converts ONNX and TensorFlow models to Snapdragon-native format.
Qualcomm Device Cloud — remote access to physical Snapdragon hardware for testing without owning devices.
ONNX Runtime Execution Provider — plugin for running ONNX models with Hexagon NPU acceleration.
Developer Discord — discord.com/invite/qualcommdevelopernetwork.

Ecosystem and Integrations

Verified integrations with Amazon SageMaker, Hugging Face, Roboflow, Dataloop, EyePop.ai, and NOTA AI for model preparation and deployment.
ExecuTorch (Meta/PyTorch) includes a stable Qualcomm AI Engine backend for deploying PyTorch models directly to the Hexagon NPU.
Llama 3.2 1B and 3B models from Meta were co-optimized for Snapdragon via QAT+LoRA quantization, achieving over 350 tokens/s prefill on Galaxy S24+.
Copilot+ PCs built on Snapdragon X Elite are Windows-native AI development targets, with NPU exposure via DirectML and QNN.

Start building with the Qualcomm AI Hub quickstart or clone the ai-hub-models repository and run pip install qai_hub_models to access the full model library.

Edit on GitHub

Qualcomm Qualcomm Snapdragon AI technology Hackathon projects

Discover innovative solutions crafted with Qualcomm Qualcomm Snapdragon AI technology, developed by our community members during our engaging hackathons.

LifePilot — On-Device Wellness Co-Pilot

Every AI wellness app works the same way: your tasks, sleep, spending, and health data get shipped to someone else's servers before you get an answer. LifePilot removes that trade-off — every model runs on the device itself. LifePilot is a calm, offline-first wellness app with four features, each backed by its own on-device model: • Overwhelm Manager — describe what's overwhelming you; an on-device Llama 3.2 1B model breaks it into a concrete 5–8 step plan, personalized from on-device memory that never syncs anywhere. • Energy Planner — a trained time-series model predicts your energy curve for the day from recent sleep and activity. • Hydration Tracker — a trained regression model sets a personalized daily water target and explains why. • Expense Scanner — point the camera at a receipt; on-device OCR plus trained models extract the merchant, total, currency, and category. No photo or text ever uploaded. Every model is exported to ExecuTorch and runs directly on the phone's own chip. There is no server-side inference path — put the phone in airplane mode and everything still works identically. That's not a fallback; it's the only mode. Where AMD fits: the four trained models (energy, hydration, and two expense models) are trained and exported on AMD Instinct MI300X GPUs via ROCm, on AMD's ROCm cloud notebooks — real GPU compute producing the exact .pte files that ship in the app. All four features are built and running on a real Android device today, including the on-device Llama agent generating multi-step task breakdowns in airplane mode. Privacy isn't a marketing line here — it's the architecture.

Hybrid Token-Efficient Routing Agent

The Hybrid Token-Efficient Routing Agent minimizes LLM inference costs by routing every query through a three-tier pipeline: an in-memory semantic cache, a local Gemma 2B model (via llama-cpp-python, designed for ROCm on AMD Instinct GPUs), and a two-tier remote escalation path through the Fireworks AI API (Gemma 2 9B IT, then Qwen 3.7 Plus for harder cases). Rather than trusting the local model's self-reported confidence, the router computes a real confidence score from the mean log-probability across generated tokens, converting it to a 0–1 scale via exponentiation. Only when this score falls below a configurable threshold (default 0.75) does the query escalate to remote inference — keeping easy queries cheap and fast while reserving paid API calls for genuinely hard cases. In benchmark testing across 8 diverse tasks (factual QA, math, sentiment, summarization, NER, code debugging, logic, code generation), the agent resolved 37.5% of queries locally at zero API cost, escalating the remainder based on task difficulty and local confidence, completing the full batch in under 6 minutes — well within the hackathon's runtime constraints. A runtime deadline guard ensures the container never exceeds the time limit regardless of query mix, falling back gracefully rather than risking a timeout.

Toddle AI

Toddle AI is an on-device Android prototype built to help parents capture and understand toddler walking patterns more clearly and privately. A parent can record a new clip or import an existing side-view walking video, and the app evaluates whether the recording is usable by checking framing, visibility, and whether enough continuous walking steps are present. Once a clip passes the quality gate, Toddle AI runs pose estimation locally on the device using a 33-landmark body model, then applies a deterministic gait-analysis pipeline to detect steps, measure temporal walking features, and generate structured observations. The app focuses on explainable metrics rather than black-box scoring, so each result comes from a measurable part of the walking sequence instead of an unsupported diagnosis. Toddle AI also includes a local AI assistant that helps explain the structured results in parent-friendly language and answer follow-up questions without sending raw video to a server. The current prototype is designed around privacy, transparency, and better home capture guidance. It is a research prototype for educational support and tracking, not a clinical diagnostic tool.

Lantern: Model anything

Lantern is an AI-powered mobile application that transforms a standard smartphone into a real-time 3D scanner using entirely on-device inference. Users simply point their phone at an object and walk around it for approximately 30 seconds while Lantern estimates depth, tracks camera motion, reconstructs geometry, and generates a dimensionally accurate 3D mesh, all without sending data to the cloud. Built with ExecuTorch, Lantern demonstrates how edge AI can deliver fast, private, and offline 3D reconstruction on consumer smartphones. By eliminating the need for expensive LiDAR hardware, external computers, or cloud processing, Lantern makes high-quality 3D scanning more accessible to engineers, designers, makers, researchers, and students. The generated 3D models can be used for reverse engineering, CAD workflows, rapid prototyping, digital asset creation, manufacturing, and AR/VR applications. Lantern showcases the power of on-device AI by combining computer vision, monocular depth estimation, and real-time 3D reconstruction into a seamless mobile experience. Our goal is to make professional-quality 3D scanning available to anyone with a smartphone while preserving user privacy and enabling low-latency performance through efficient edge inference.

Memory Recall Assistant

Memory Recall Assistant is a fully on-device, multimodal AI memory aid designed for people with early-to-mid stage cognitive decline. The app runs continuously in the background, periodically capturing context through the camera and microphone. When the user presses Recall while looking at someone, the app identifies the person using CavaFace face embeddings (w8a16 quantized, ~99.5% accuracy on LFW benchmark), retrieves stored conversation history transcribed by Whisper-Small-Quantized, and generates a warm, natural memory cue using Llama 3.2 1B Instruct - all within seconds and entirely offline. A key design innovation is center crop targeting: rather than identifying the largest or highest-confidence face in frame, the app identifies the face closest to the center of the camera frame - the person the user is actively talking to, not a bystander. This makes identification intentional and natural, and extends directly to Meta AI Glasses where the camera points where the user looks. All sensitive data - face embeddings, conversation transcripts, location context - is stored on-device in a local SQLite knowledge graph and never transmitted anywhere. The system functions fully in airplane mode, demonstrating that privacy-sensitive AI for vulnerable populations is not only possible on mobile hardware - it is better this way. There is no server to breach, no API to call, and no company holding your family's data.

SafeScreen AI

SafeScreen AI is designed as a local first line of defense. It runs directly on a Snapdragon-powered Android device using ExecuTorch and analyzes visual content on-device in real time. When the system detects potentially explicit, abusive, or manipulated media, it can immediately warn, blur, redact, mask, or block the content directly on screen before the user fully engages with it. Our initial focus is on two high-impact use cases: 1. Real-time explicit and harmful visual content protection, especially for young kids, teens, and women who may be targeted by unsafe content, online abuse, harassment, impersonation, or AI-generated explicit media. Many vulnerable users may not have the technical awareness to recognize manipulated or harmful content before it affects them. SafeScreen AI acts as a private safety shield that can blur or redact harmful content in the moment, reducing exposure before harm spreads. 2. Deepfake and manipulated media detection, helping users recognize synthetic or altered images and videos before trusting, sharing, or being harmed by them. This matters because AI-generated abuse is no longer hypothetical. Recent reporting has found AI-generated explicit deepfakes spreading in schools and affecting hundreds of students globally. Research on publicly available deepfake model variants found nearly 35,000 downloadable deepfake-related models, with almost 15 million downloads since late 2022; 96% of the targeted individuals were women. Online harm is also increasingly recognized by governments and advocacy groups as a serious form of non-consensual intimate image abuse. Using ExecuTorch and Snapdragon acceleration, we aim to build a low-latency, privacy-preserving pipeline that captures visual input, runs lightweight models locally, and responds immediately with protective actions such as warning, blurring, redaction, masking, or blocking.