Google Open Source

@GoogleOSS

Announcing new open source releases, exploring projects, sharing how we approach FOSS, and supporting communities around the world.

Mountain View, CA opensource.google Joined March 2017

145 Following 82K Followers 2.5K Posts

Latest posts 5

Pinned post

Google Open Source @GoogleOSS 2d

With 73M+ monthly PyPI downloads, kubernetes-client/python powers AI/ML from Kubeflow to Ray. How Google Cloud & @kubernetesio SIG API Machinery are modernizing it for hyperscale clusters—which feature matters most? 👇 Link in first reply

Dark graphic card 1 of 4 for Google Open Source Blog titled "Overhauling the Official Kubernetes Python Client for the AI/ML Era" with subtitle "73M+ monthly PyPI downloads powering Kubeflow, Ray, KServe, and Airflow." Center rounded panel displays the Google Open Source Kubernetes circuit-board graphic (blue Kubernetes wheel surrounded by code, cloud, gear, and community icons) above attribution for "Yu Liao, Haowei Cai & Federico Bongiovanni — Google Cloud Kubernetes Engineering & SIG API Machinery" and two bullets on 50,000-node AI/ML clusters and OpenAPI Generator v7.20+, asyncio, SharedInformer, and HTTP/2. Footer bar: "Swipe for architecture details →" and "opensource.googleblog.com • goo.gle/k8s-python-overhaul".

Dark graphic card 2 of 4 titled "Why Hyperscale AI/ML Needs a Modern Python Client" with header badge "SCALE & BOTTLENECKS". Three color-accented metric cards list: (1) Ecosystem Scale (Blue): "73M+ PyPI Downloads / Month — ~1.7M Daily Pulls & 7,500+ GitHub Stars. Core bridge for Kubeflow, Ray, KServe & Airflow—still running OpenAPI Generator v4.3.0 (removed upstream in 2022)." (2) Large-Cluster Latency (Yellow): "30+ Seconds to List Cluster Pods — Model Reconfiguration & HTTP/1.1 Limits. Repeated JSON schema reconfiguration stalls pod queries while HTTP/1.1 and sync-only I/O bottleneck high-concurrency workloads." (3) Engineering Focus (Green): "Up to 50,000-Node AI/ML Fleets — Targeted Primitives Over 1:1 Go Parity. Google Cloud and SIG API Machinery prioritize high-ROI Python gaps rather than replicating every long-tail client-go feature."

Dark graphic card 3 of 4 titled "Three Pillars Behind the Modernization Roadmap" with a 2x2 grid of technical roadmap cards: (1) Pillar 1 // Machinery (Blue): "OpenAPI v7.20+ — Replaces dead v4.3.0 templates (~400K lines) with modern type hints and clean ML pip deps" (kubernetes-client/gen). (2) Pillar 2 // Concurrency (Green): "asyncio & Informers — Official async I/O plus SharedInformer list-watch caching for TrainJob and RayCluster" (python PR #2515). (3) Pillar 2 // Streaming (Yellow): "WebSocket v5 — Transitions exec, attach & port-forward off legacy SPDY to v5.channel.k8s.io (KEP-4006)" (python PR #2486). (4) Pillar 3 // Performance (Red): "Fast JSON & HTTP/2 — Cuts model init overhead during cluster queries and evaluates HTTP/2 multiplexing" (issue #1921).

Dark graphic card 4 of 4 for X (@GoogleOSS) titled "Built in the Open With SIG API Machinery". Left box "UPSTREAM REPOSITORIES & HANDLES" lists "@kubernetesio • @CloudNativeFdn", "github.com/kubernetes-client/python & /gen", "kubernetes.slack.com (#sig-api-machinery)", and "Open bi-weekly triage meetings & PRs (#2515, #2486)". Right box "2-POST THREAD ON @GOOGLEOSS" shows "1. Post #1 (Main Tweet): 4-Card Carousel + @kubernetesio (0 links in Post #1 — link in first reply ↓)" and "2. Post #2 (First Reply): https://goo.gle/k8s-python-overhaul #Kubernetes #Python #OpenSource #CloudNative". Bottom blue CTA box: "Read the Roadmap & Join Upstream Triage — First reply below: https://goo.gle/k8s-python-overhaul".

Google Open Source @GoogleOSS 1d

What are your open source weekend plans?

Google Open Source @GoogleOSS 1d

TWiOS #22 is out.

From Google donating gVisor to @CloudNativeFdn and @openssf security lessons to CI-testing Agent Skills, ATProto Spaces, and open FOSDEM & CNCF EU CFPs.

👇 Link in first reply

Google Open Source @GoogleOSS 1d

Design with an elephant. Review with a goldfish.

Watch the 65-second @googledevs explainer on how the open source Elephant-Goldfish Model catches blind spots before AI agents write code.

👇 Video and blog post in first reply

Dark-background title card for the 65-second Elephant-Goldfish Model (EGM) explainer video. At the top center is the minimalist split EGM mark: a blue elephant silhouette on the left facing an orange goldfish silhouette on the right. Centered white text below reads "Elephant-Goldfish Model" above the tagline "Design with an elephant. Review with a goldfish." ("elephant" in blue, "goldfish" in orange), with a bottom badge reading "Open source, from Google · egm.dev".

Google Open Source @GoogleOSS 3d

A2UI 1.0 is out! 🎨

An Apache 2.0 open protocol for generative UI: agents send declarative data, and @FlutterDev, React, Angular, Lit & SwiftUI render it natively.

With @GoogleCloudTech & @googledevs.
👇 Links in first reply