All terms
Frameworks & Tools
ONNX Runtime
A cross-platform engine for running and accelerating models in ONNX format.
Definition
ONNX Runtime is an engine for running models stored in the ONNX format. It executes them across operating systems and hardware, using backend-specific optimizations and accelerators to speed up inference. By separating the model format from the runtime, it lets a single exported model be deployed efficiently on CPUs, GPUs, and specialized chips.