OpenAI models land on Google Cloud, putting GPT-6 on all three major US clouds

CNBC frames the move as a further expansion beyond Microsoft. Google Cloud customers can now call OpenAI's frontier models directly from Model Garden, alongside Gemini 4 Argon, which Google is offering at introductory pricing of $2/$10 per million input/output tokens until December 1.

Mechanically, this is a distribution deal: OpenAI gains enterprise reach and procurement convenience, while Google gains a reason for customers to keep workloads on GCP even when they prefer GPT. It mirrors AWS's Bedrock Managed Agents, which are built on a customized OpenAI Agents API.

Competitively, Azure exclusivity is effectively over as a selling point. HN readers (239 points) were amused by disclaimers reaffirming the Microsoft partnership and read the deal as hyperscaler hedging, while debating whether wider distribution raises distillation risk.

Watch for pricing parity across clouds, whether Microsoft responds with deeper Copilot bundling, and whether Anthropic and Google models follow a similar everywhere strategy.