On November 3, 2026, OpenAI revealed that GPT-6 Astra designed an entire hardware device for the Codex coding agent in under a week: enclosure, circuit board, firmware, and interface. The demo measures a gap that is still wide between AI and real physical engineering, worth reading before you wire your design flows to a model.

On October 21, 2026, Google Cloud put OpenAI’s models — GPT-6 Astra, GPT-6.1 Sol and GPT-6 Luna — into its Model Garden, becoming the only hyperscaler offering Gemini, Claude and GPT on a single platform. For a CISO the story is not the catalog but the governance: these models are billed through Marketplace, stay inside your data perimeter and plug into VPC Service Controls.

On September 22, 2026, Anthropic shipped Claude Opus 5.5 with a 20% price cut, and OpenAI answered minutes later with two GPT-6 models, Sol and Luna, priced at half their predecessors. Model choice is now a budget decision as much as a technical one: benchmark cost per token against your real workload.