NVIDIA Jetson

LiteRT

google-ai-edge/LiteRT

Popular

Google's on-device ML runtime formerly called TensorFlow Lite, for Android, iOS, embedded Linux and accelerators.

TensorFlow Lite Micro

tensorflow/tflite-micro

Popular

LiteRT for Microcontrollers: tiny interpreter running quantized models in kilobytes of RAM; the classic TinyML runtime.

ExecuTorch

pytorch/executorch

Trending

PyTorch's on-device runtime (1.0 GA in 2025) for phones, wearables and MCUs with Arm Ethos-U and NPU backends.

ONNX Runtime

microsoft/onnxruntime

Trending (detected automatically)Popular

Cross-platform inference accelerator for ONNX models on CPU, GPU, NPU, mobile and the web.

Edge Impulse

edgeimpulse.com

Popular

End-to-end platform to collect sensor data, train and deploy optimized models to MCUs; now part of Qualcomm.

llama.cpp

ggml-org/llama.cpp

Popular

C/C++ LLM inference engine behind GGUF and most local tools; runs on CPUs, Apple Silicon and GPUs.

Apache TVM

apache/tvm

Open deep learning compiler that optimizes models for diverse CPUs, GPUs and accelerators, including edge targets.

emlearn

emlearn/emlearn

Converts scikit-learn and Keras models into portable, dependency-free C code for microcontrollers.