At the 2025 PyTorch Conference, the PyTorch Foundation announced several initiatives aimed at advancing open, scalable AI infrastructure. The foundation welcomed Ray, the distributed computing framework, as a hosted project and introduced PyTorch Monarch, a new framework simplifying distributed AI workload across multiple machines. The event also spotlighted new open research projects, including Stanford’s Marin and AI2’s Olmo-Thinking, highlighting the growing push for transparency and reproducibility in foundation model development.
The inclusion of Ray reflects the foundation’s broader strategy to build a unified open ecosystem spanning model development, serving, and distributed execution. Originally developed at UC Berkeley’s RISELab, Ray provides a compact set of Python primitives that make distributed computation as intuitive as writing local code, enabling developers to scale training, tuning, and inference workloads seamlessly.
The addition of Ray complements other recent projects under the foundation’s umbrella, including DeepSpeed for distributed training and vLLM for high-throughput inference. Together, PyTorch, DeepSpeed, vLLM, and Ray form a cohesive open-source stack covering the full model lifecycle—from experimentation to production-scale deployment.
In parallel, the Meta PyTorch team introduced PyTorch Monarch, a framework designed to abstract entire GPU clusters as a single logical device. Monarch’s array-like mesh interface allows developers to express parallelism using Pythonic constructs while the system automatically manages data and computation distribution. Built on a Rust-based backend, Monarch aims to combine performance with safety and reduce the cognitive load of distributed programming.
The conference further emphasized open collaboration in foundation model development and research. In a keynote, Percy Liang from Stanford University introduced Marin, an open lab under the Center for Research on Foundation Models that seeks to make frontier AI development fully transparent—releasing datasets, code, hyperparameters, and training logs to enable reproducibility and community participation.
Similarly, Nathan Lambert, Senior Research Scientist from Ai2, presented Olmo-Thinking, an open reasoning model that disclosed details about the training process, model architecture decisions, data sourcing, and training code design often absent in the closed model releases. These efforts align with a broader movement toward open, and reproducible foundation models.
By expanding its scope beyond core framework development, the PyTorch Foundation is positioning itself as a central hub for open AI infrastructure. The upcoming 2026 PyTorch Conference in San Jose will likely continue this focus on ecosystem collaboration and developer enablement.
