phyai

PhyAI is a high-performance framework for running Physical AI models (VLA, WAM, and beyond), supporting both cloud-based serving and on-device deployment.

What is phyai?

PhyAI is a high-performance framework for running Physical AI models (VLA, WAM, and beyond), supporting both cloud-based serving and on-device deployment.

Overview of phyai

What it is

PhyAI is a latency-first serving engine for Physical AI. It is designed for latency-critical workloads, such as policy and action models that run in interactive systems. The framework supports both cloud-based serving and on-device deployment, and it can run on NVIDIA Jetson edge devices as well as scale to GPU clusters.

Key features

  • Runs on NVIDIA Jetson edge devices - Scales to GPU clusters with DP, TP, and CFG parallelism - Uses high-performance kernels from FlashInfer and Humming - Supports INT4, INT8, FP8, MXFP4, and NVFP4 quantization through kernel policies (PR under review)

Use cases

PhyAI is intended for running Physical AI models, including Vision-Language-Action (VLA) models, Whole-body Action Models (WAM), and foundation models. It is suited for latency-critical interactive systems such as robotics policy and action inference. The engine supports both single-host and multi-host serving, enabling deployment on edge devices or across GPU clusters.

Getting started

From source: bash git clone https://github.com/mingti-org/phyai cd phyai uv sync Nightly build: bash uv pip install phyai phyai-ext --extra-index-url https://mingti-org.github.io/phyai/simple/ --prerelease=allow For detailed installation instructions, see the PhyAI installation guide.

License and pricing

PhyAI is released under the MIT License. It is open source with no stated pricing.

Based on the project's public README and repository metadata.

Add this badge to your README

Featured on YunoDB Badge styles & help

Maintain this project? Show that it is featured on YunoDB.

Markdown[![Featured on YunoDB](https://yunodb.com/badge/mingti-orgphyai.svg)](https://yunodb.com/tools/mingti-orgphyai?utm_source=github&utm_medium=badge)
HTML<a href="https://yunodb.com/tools/mingti-orgphyai?utm_source=github&utm_medium=badge"><img src="https://yunodb.com/badge/mingti-orgphyai.svg" alt="Featured on YunoDB" height="20"></a>