Ray

Data Frameworks

Ray is an open-source unified compute framework for AI and Python applications, originally developed at UC Berkeley's RISELab and now maintained by the Ray Project under the Linux Foundation.

Reviewed by 7wData
Share Email

On this page

Profile

Ray is an open-source framework that distributes Python and AI workloads across clusters, enabling scalable training, serving, and data processing.

Ray is an open-source unified compute framework for AI and Python applications, originally developed at UC Berkeley's RISELab and now maintained by the Ray Project under the Linux Foundation. The project provides a distributed execution engine for scaling machine learning, reinforcement learning, and general Python workloads from a single laptop to a large cluster. As of June 2026, the Ray GitHub repository has over 42,800 stars, 7,600 forks, and more than 30,600 commits, with active contributions from engineers at Anyscale, Google, and other organizations.

The core product is the Ray runtime, which includes libraries for reinforcement learning (RLlib), hyperparameter tuning (Tune), data processing (Datasets), model serving (Serve), and workflow orchestration. Ray is used by major technology companies and AI labs for training and serving large models, powering production systems at OpenAI, Uber, Netflix, and others. The project is backed by Anyscale, the commercial entity founded by the original Berkeley researchers, which raised $100 million in Series C funding in 2022 at a valuation of over $1 billion.

Ray has no disclosed standalone revenue or funding as an open-source project; its financial trajectory is tied to Anyscale's commercial offerings. Recent development activity includes support for Google TPUs, integration with vLLM 0.21.0 for LLM serving, and automated dependency upgrades via Claude AI skills. The project's governance remains community-driven, with code reviews and contributions from multiple organizations, though Anyscale employees account for a significant share of commits.

Track Ray and 240+ vendors.

335k+ subscribers read the daily AI & data note. One email, both newsletters. Unsubscribe anytime.

Who buys this

  • AI research labs and universities running distributed reinforcement learning and hyperparameter tuning experiments
  • Enterprise machine learning teams deploying and serving large language models in production
  • Data science and engineering teams processing large-scale datasets with Python
  • Cloud service providers offering managed Ray services (e.g., Google Cloud, AWS, Azure)
  • Startups and mid-market companies building AI-powered applications that require scalable compute

Strengths and what to watch

Strengths

  • Widely adopted open-source project with over 42,800 GitHub stars and contributions from multiple organizations, indicating strong community trust and ecosystem
  • Covers the full ML lifecycle from training to serving with integrated libraries (RLlib, Tune, Serve, Datasets), reducing the need for multiple tools
  • Active development with recent support for Google TPUs and vLLM integration, keeping pace with hardware and model serving advancements

Watch for

  • Heavy reliance on Anyscale, the commercial entity founded by original Berkeley researchers, for core maintenance and direction; governance is community-driven but Anyscale employees dominate commits
  • Competition from alternative distributed compute frameworks like Dask, Modin, and Spark, as well as managed services from cloud providers that may reduce Ray's differentiation
  • No disclosed revenue or funding for the open-source project itself; financial sustainability depends on Anyscale's commercial success, which is not publicly detailed

Recent moves

Key Information

Industry
Data Frameworks
Founded
1986

Sources

  1. github.com — GitHub repository stats (42.8k stars, 7.6k forks, 30,602 commits), recent commits, and contributor activity
  2. pitchbook.com — PitchBook profile for a different company named Project Ray (smartphone for visually impaired), not the open-source Ray project; included as a cautionary note about name collision
  3. www.cbinsights.com — CB Insights profile for Ray Therapeutics (biotech), not the open-source Ray project; included as a cautionary note about name collision
  4. ai2roi.substack.com — Newsletter discussing OpenAI's $122B round and enterprise AI trends; no direct mention of Ray project
  5. www.cnbc.com — CNBC article on Meta layoffs and AI investment; no direct mention of Ray project
  6. news.crunchbase.com — Crunchbase tech layoffs tracker; no direct mention of Ray project
  7. www.lhh.com — LHH research on layoff trends; no direct mention of Ray project