{"data":{"slug":"ray-project-ray","name":"ray","tagline":"Ray is an AI compute engine with a core distributed runtime and AI Libraries for accelerating ML workloads.","github_url":"https://github.com/ray-project/ray","owner":"ray-project","repo":"ray","owner_avatar_url":"https://avatars.githubusercontent.com/u/22125274?v=4","primary_language":"Python","stars":43526,"forks":7929,"topics":["data-science","deep-learning","deployment","distributed","hyperparameter-optimization","hyperparameter-search","large-language-models","llm","llm-inference","llm-serving","machine-learning","optimization","parallel","python","pytorch","ray","reinforcement-learning","rllib","serving","tensorflow"],"archived":false,"github_pushed_at":"2026-08-16T00:26:16+00:00","maintenance_label":"Very active","stars_delta_30d":270,"url":"https://www.graphcanon.com/tools/ray-project-ray","markdown_url":"https://www.graphcanon.com/tools/ray-project-ray.md","api_url":"https://www.graphcanon.com/api/graphcanon/tools/ray-project-ray","graph_url":"https://www.graphcanon.com/api/graphcanon/graph?tool=ray-project-ray","description":"Ray is an AI compute engine. Ray consists of a core distributed runtime and a set of AI Libraries for accelerating ML workloads.","homepage_url":"https://ray.io","license":"Apache-2.0","open_issues":3483,"watchers":479,"ai_summary":"A framework designed to simplify the process of writing applications that run across many machines. It includes various libraries like RLlib (for reinforcement learning) and support for large language models serving and inference.","readme_excerpt":".. image:: https://github.com/ray-project/ray/raw/master/doc/source/images/ray_header_logo.png\n\n.. image:: https://readthedocs.org/projects/ray/badge/?version=master\n    :target: http://docs.ray.io/en/master/?badge=master\n\n.. image:: https://img.shields.io/badge/Ray-Join%20Slack-blue\n    :target: https://www.ray.io/join-slack\n\n.. image:: https://img.shields.io/badge/Discuss-Ask%20Questions-blue\n    :target: https://discuss.ray.io/\n\n.. image:: https://img.shields.io/twitter/follow/raydistributed.svg?style=social&logo=twitter\n    :target: https://x.com/raydistributed\n\n.. image:: https://img.shields.io/badge/Get_started_for_free-3C8AE9?logo=data%3Aimage%2Fpng%3Bbase64%2CiVBORw0KGgoAAAANSUhEUgAAABAAAAAQCAYAAAAf8%2F9hAAAAAXNSR0IArs4c6QAAAERlWElmTU0AKgAAAAgAAYdpAAQAAAABAAAAGgAAAAAAA6ABAAMAAAABAAEAAKACAAQAAAABAAAAEKADAAQAAAABAAAAEAAAAAA0VXHyAAABKElEQVQ4Ea2TvWoCQRRGnWCVWChIIlikC9hpJdikSbGgaONbpAoY8gKBdAGfwkfwKQypLQ1sEGyMYhN1Pd%2B6A8PqwBZeOHt%2FvsvMnd3ZXBRFPQjBZ9K6OY8ZxF%2B0IYw9PW3qz8aY6lk92bZ%2BVqSI3oC9T7%2FyCVnrF1ngj93us%2B540sf5BrCDfw9b6jJ5lx%2FyjtGKBBXc3cnqx0INN4ImbI%2Bl%2BPnI8zWfFEr4chLLrWHCp9OO9j19Kbc91HX0zzzBO8EbLK2Iv4ZvNO3is3h6jb%2BCwO0iL8AaWqB7ILPTxq3kDypqvBuYuwswqo6wgYJbT8XxBPZ8KS1TepkFdC79TAHHce%2F7LbVioi3wEfTpmeKtPRGEeoldSP%2FOeoEftpP4BRbgXrYZefsAI%2BP9JU7ImyEAAAAASUVORK5CYII%3D\n   :target: https://www.anyscale.com/ray-on-anyscale?utm_source=github&utm_medium=ray_readme&utm_campaign=get_started_badge\n\nRay is a unified framework for scaling AI and Python applications. Ray consists of a core distributed runtime and a set of AI libraries for simplifying ML compute:\n\n.. image:: https://github.com/ray-project/ray/raw/master/doc/source/images/what-is-ray-padded.svg\n\n..\n  https://docs.google.com/drawings/d/1Pl8aCYOsZCo61cmp57c7Sja6HhIygGCvSZLi_AuBuqo/edit\n\nLearn more about `Ray AI Libraries`_:\n\n- `Data`_: Scalable Datasets for ML\n- `Train`_: Distributed Training\n- `Tune`_: Scalable Hyperparameter Tuning\n- `RLlib`_: Scalable Reinforcement Learning\n- `Serve`_: Scalable and Programmable Serving\n\nOr more about `Ray Core`_ and its key abstractions:\n\n- `Tasks`_: Stateless functions executed in the cluster.\n- `Actors`_: Stateful worker processes created in the cluster.\n- `Objects`_: Immutable values accessible across the cluster.\n\nLearn more about Monitoring and Debugging:\n\n- Monitor Ray apps and clusters with the `Ray Dashboard <https://docs.ray.io/en/latest/ray-core/ray-dashboard.html>`__.\n- Debug Ray apps with the `Ray Distributed Debugger <https://docs.ray.io/en/latest/ray-observability/ray-distributed-debugger.html>`__.\n\nRay runs on any machine, cluster, cloud provider, and Kubernetes, and features a growing\n`ecosystem of community integrations`_.\n\nInstall Ray with: ``pip install ray``. For nightly wheels, see the\n`Installation page <https://docs.ray.io/en/latest/ray-overview/installation.html>`__.\n\n.. _`Serve`: https://docs.ray.io/en/latest/serve/index.html\n.. _`Data`: https://docs.ray.io/en/latest/data/data.html\n.. _`Workflow`: https://docs.ray.io/en/latest/workflows/\n.. _`Train`: https://docs.ray.io/en/latest/train/train.html\n.. _`Tune`: https://docs.ray.io/en/latest/tune/index.html\n.. _`RLlib`: https://docs.ray.io/en/latest/rllib/index.html\n.. _`ecosystem of community integrations`: https://docs.ray.io/en/latest/ray-overview/ray-libraries.html\n\n\nWhy Ray?\n--------\n\nToday's ML workloads are increasingly compute-intensive. As convenient as they are, single-node development environments such as your laptop cannot scale to meet these demands.\n\nRay is a unified way to scale Python and AI applications from a laptop to a cluster.\n\nWith Ray, you can seamlessly scale the same code from a laptop to a cluster. Ray is designed to be general-purpose, meaning that it can performantly run any kind of workload. If your application is written in Python, you can scale it with Ray, no other infrastructure required.\n\nMore Information\n----------------\n\n- `Documentation`_\n- `Ray Architecture whitepaper`_\n- `Exoshuffle: large-scale data shuffle in Ray`","github_created_at":"2016-10-25T19:38:30+00:00","created_at":"2026-07-07T17:31:29.637319+00:00","updated_at":"2026-08-16T12:01:38.796179+00:00","categories":[{"slug":"inference-serving","name":"Inference & Serving","url":"https://www.graphcanon.com/categories/inference-serving","markdown_url":"https://www.graphcanon.com/categories/inference-serving.md","api_url":"https://www.graphcanon.com/api/graphcanon/categories/inference-serving"},{"slug":"model-training","name":"Model Training","url":"https://www.graphcanon.com/categories/model-training","markdown_url":"https://www.graphcanon.com/categories/model-training.md","api_url":"https://www.graphcanon.com/api/graphcanon/categories/model-training"}],"tags":[{"slug":"data-science","name":"data-science"},{"slug":"deep-learning","name":"deep-learning"},{"slug":"deployment","name":"deployment"},{"slug":"distributed","name":"distributed"},{"slug":"hyperparameter-optimization","name":"hyperparameter-optimization"},{"slug":"large-language-models","name":"large language models"},{"slug":"llm-inference","name":"llm-inference"},{"slug":"machine-learning","name":"machine-learning"}],"trust":{"provenance":{"is_fork":false,"github_id":71932349,"owner_type":"Organization","methodology":"github_public_v1","parent_repo":null,"near_duplicate_slugs":[]},"computed_at":"2026-08-16T12:01:38.039Z","maintenance":{"label":"Very active","score":96,"methodology":"github_public_v1","releases_90d":3,"days_since_push":0,"last_release_at":"2026-08-11T01:15:44Z","stars_delta_30d":270,"open_issues_delta_30d":14},"security_summary":{"status":"no_lockfile","scanner":null,"low_count":0,"high_count":0,"last_scan_at":"2026-07-11T10:57:22.165Z","medium_count":0,"scan_profile":"none","critical_count":0}},"capability_facts":{"scan":{"source":"repo_scan","observed_at":"2026-08-16T12:01:38.468Z"},"languages":{"value":["python"],"source":"github.language+pyproject.toml","observed_at":"2026-08-16T12:01:38.468Z"},"license_spdx":{"value":"Apache-2.0","source":"github.license","observed_at":"2026-08-16T12:01:38.468Z"}},"decision_facts":{"hosting":null,"pricing":null,"requirements":{"min_ram_gb":null,"requires_docker":false},"constraints":{"min_ram_gb":null,"requires_docker":false},"when_to_use":["When you need to develop applications that require the distribution of tasks across multiple machines.","For projects involving reinforcement learning, thanks to Ray's RLlib library specifically tailored for these needs.","If deploying large language models and focusing on both training and inference stages where a unified ecosystem benefits performance and ease-of-use.","When hyperparameter optimization is required; Ray Tune within the Ray framework can provide efficient search strategies."],"when_not_to_use":["For simplistic projects or single-machine use cases, as Ray's distributed architecture may introduce unnecessary complexity.","If your project strictly adheres to languages other than Python, since most of the ecosystem and support revolves around Python.","When an environment already heavily utilizes another distributed computing framework that integrates deeply with specific needs, moving to Ray might not offer additional advantages over sticking with,","for example,\nan existing, well-integrated solution like Apache Spark for data processing."],"source":"enrich:decision_facts","observed_at":"2026-07-11T12:33:15.176Z"},"constraint_facets":{"min_ram_gb":null,"requires_docker":false},"decision_summary":[{"label":"Adopt for","value":"Ray offers a core distributed runtime and specialized libraries for optimizing ML workloads in Python."},{"label":"License detail","value":"Apache-2.0 license allows for both commercial and private use without the need to open-source your entire project."}]}}