Qualcomm® AI Hub Models

The Qualcomm® AI Hub Models are a collection of state-of-the-art machine learning models optimized for deployment on Qualcomm® devices.

• List of Models by Category
• On-Device Performance Data
• Device-Native Sample Apps

See supported: On-Device Runtimes, Hardware Targets & Precision, Chipsets, Devices

 

NEW: Quick Start with the AI Hub Models CLI

Use our lightweight command-line interface to browse and download from the collection of Qualcomm® AI Hub Models.

pip install qai_hub_models_cli # (the CLI is also available with the qai-hub-models package)

qai-hub-models models                                                # browse the catalog
qai-hub-models info mobilenet_v2                                     # model details + download options
qai-hub-models fetch mobilenet_v2 --runtime tflite --precision float # download a deployable asset
# ... and more

The CLI also offers a Python API.

See the CLI README for full usage instructions.

 

Setup

1. Install Python Package

The package is available via pip:

# NOTE for Snapdragon X Elite and Snapdragon X2 Elite users:
# Only AMDx64 (64-bit) Python is supported on Windows.
# Installation will fail when using Windows ARM64 Python.

pip install qai_hub_models

Some models (e.g. YOLOv7) require additional dependencies. View the model README (at qai_hub_models/models/model_id) for installation instructions.

 

2. Configure AI Hub Workbench Access

Many features of AI Hub Models (such as model compilation, on-device profiling, etc.) require access to Qualcomm® AI Hub Workbench:

• Create a Qualcomm® ID, and use it to login to Qualcomm® AI Hub Workbench.
• Configure your API token: qai-hub configure --api_token API_TOKEN

 

Getting Started

Export and Run A Model on a Physical Device

All models in our directory can be compiled and profiled on a hosted Qualcomm® device:

pip install "qai_hub_models[yolov7]"

qai-hub-models export yolov7 --target-runtime tflite --precision float --device "Samsung Galaxy S25 (Family)"

Using Qualcomm® AI Hub Workbench, the export script will:

1. Compile the model for the chosen device and target runtime (see: Compiling Models on AI Hub Workbench).
2. If applicable, Quantize the model (see: Quantization on AI Hub Workbench)
3. Profile the compiled model on a real device in the cloud (see: Profiling Models on AI Hub Workbench).
4. Run inference with a sample input data on a real device in the cloud, and compare on-device model output with PyTorch output (see: Running Inference on AI Hub Workbench)
5. Download the compiled model to disk.

 

End-To-End Model Demos

Most models in our directory contain CLI demos that run the model end-to-end:

pip install "qai_hub_models[yolov7]"
# Predict and draw bounding boxes on the provided image
python -m qai_hub_models.models.yolov7.demo [--image ...] [--eval-mode {fp,on-device}] [--help]

End-to-end demos:

1. Preprocess human-readable input into model input
2. Run model inference
3. Postprocess model output to a human-readable format

Many end-to-end demos use AI Hub Workbench to run inference on a real cloud-hosted device (with --eval-mode on-device). All end-to-end demos can also run locally via PyTorch (with --eval-mode fp).

 

Sample Applications

Native applications that can run our models (with pre- and post-processing) on physical devices are published in the AI Hub Apps repository.

Python applications are defined for all models (from qai_hub_models.models.<model_name> import App). These apps wrap model inference with pre- and post-processing steps written using torch & numpy. These apps are optimized to be an easy-to-follow example, rather than to minimize prediction time.

 

Model Support Data

On-Device Runtimes

| Runtime | Supported OS | | -- | -- | | Qualcomm AI Engine Direct | Android, Linux, Windows | LiteRT (TensorFlow Lite) | Android, Linux | ONNX | Android, Linux, Windows

Device Hardware & Precision

| Device Compute Unit | Supported Precision | | -- | -- | | CPU | FP32, INT16, INT8 | GPU | FP32, FP16 | NPU (includes Hexagon DSP, HTP) | FP16*, INT16, INT8

*Some older chipsets do not support fp16 inference on their NPU.

Chipsets

• Snapdragon 8 Elite Gen 5, 8 Elite, 8 Gen 3, 8 Gen 2, and 8 Gen 1 Mobile Platforms
• Snapdragon X2 Elite, Snapdragon X Elite Compute Platforms
• SA7255P, SA8295P, and SA8775P Automotive Platforms
• QCS 6490, QCS 8250, QCS 9075, and QCS 8550 IoT Platforms
• QCS8450 XR Platform

and many more.

Devices

• Samsung Galaxy S21, S22, S23, S24, and S25 Series
• Xiaomi 12, 13, 15, and 17
• Snapdragon X Elite CRD and Snapdragon X2 Elite CRD (Compute Reference Device)
• Qualcomm RB3 Gen 2, RB5 Gen 2, IQ-8, IQ-9

and many more.

 

Model Directory

Computer Vision

| Model | README | | -- | -- | | | | | Image Classification | Beit | qai_hub_models.models.beit | | ConvNext-Base | qai_hub_models.models.convnext_base | | ConvNext-Tiny | qai_hub_models.models.convnext_tiny | | DLA-102-X | qai_hub_models.models.dla102x | | DenseNet-121 | qai_hub_models.models.densenet121 | | EfficientFormer | qai_hub_models.models.efficientformer | | EfficientNet-B0 | qai_hub_models.models.efficientnet_b0 | | EfficientNet-B4 | qai_hub_models.models.efficientnet_b4 | | EfficientNet-Lite4 | qai_hub_models.models.efficientnet_lite4 | | EfficientNet-V2-s | qai_hub_models.models.efficientnet_v2_s | | EfficientViT-b2-cls | qai_hub_models.models.efficientvit_b2_cls | | EfficientViT-l2-cls | qai_hub_models.models.efficientvit_l2_cls | | GPUNet | qai_hub_models.models.gpunet | | GoogLeNet |