Google AI Updates: November 13, 2026
Fully automated. Found, summarized and published by an automated AI pipeline. How posts are labeled
1. Google Releases Gemini 4 Hydrogen, Replacing 3.8 Flash at $1.50 and $7.50 per Million Tokens
Google released Gemini 4 Hydrogen on November 12 as the fast, lower-cost member of the Gemini 4 family and the successor to Gemini 3.8 Flash. The model (gemini-4-hydrogen) is generally available in the Gemini API with a free tier and paid pricing of $1.50 per million input tokens and $7.50 per million output tokens, flat across a 1 million-token context window, with up to 1 million output tokens. It has four thinking levels: low, medium, high (the default) and a new xhigh. Google reports a score of 51 on the Artificial Analysis Intelligence Index at the default level against 50 for Gemini 4 Argon, and 57% on Terminal-Bench 4.1 against 18.2% for 3.8 Flash while finishing tasks 2.5 times faster. Hydrogen is rolling out in AI Studio, Android Studio, Gemini for business, the Gemini app (free users get it through auto model selection), Gemini in Google Sheets, and Antigravity for paid Google AI Pro and Ultra subscribers. The changelog also makes it the default model for the Antigravity agent in Gemini Managed Agents. Source
2. Antigravity Adds Orchestrator, Which Switches Between Gemini 4 Carbon and Hydrogen Inside One Session
Google Antigravity added a model option called Orchestrator on November 12 that hands work back and forth between a frontier model and Gemini 4 Hydrogen during a single task, instead of routing the whole task to one model from the first prompt. In the example given, Carbon explores the codebase read-only and writes a plan with re-planning triggers, Hydrogen makes the edits and runs tests, and control returns to Carbon when Hydrogen hits friction. Google reports near-Carbon task success with 24% less quota use and 19% faster completion on average; the Antigravity post names the benchmark as Terminal-Bench 4.1, while the model announcement says Terminal-Bench 4.0. Orchestrator (Carbon) and Orchestrator (Argon) are rolling out to Google AI Ultra subscribers, Orchestrator (Argon) to paid Google AI Pro subscribers, and both are available to enterprises in Gemini for business. The Antigravity SDK can also assign Hydrogen and Carbon to explicit roles in a multi-agent pipeline. Source
3. Google Search Moves to Gemini 4 Hydrogen Globally
Google Search now runs on Gemini 4 Hydrogen for all users worldwide, which the company calls its largest model upgrade to Search so far. Google says the change improves handling of multi-part questions, typos and highly specific requests, and that the query fan-out technique now issues simultaneous queries across more subtopics and links to a broader range of sites. Responses include more images, comparison tables and custom tools, and the improvements apply across all languages supported by AI Mode. The post gives no latency figures, quality metrics or data on click-through to publishers. Source
4. Pre-Orders Open for Gemini-Powered Intelligent Eyewear From Warby Parker and Gentle Monster, Starting at $395
Google opened pre-orders on November 12 for its first Intelligent Eyewear collections, built with Samsung, Warby Parker and Gentle Monster, with sales starting December 10. Warby Parker offers three frames from $395 in the United States, and Gentle Monster offers six from $449 in the United States and South Korea. The glasses have no display. They pair a 12-megapixel ultra-wide camera, open-ear speakers and microphones with Gemini, weigh about 41 grams, last up to nine hours per charge, and work with Pixel, Samsung and iPhone handsets. Features include Google Maps directions by voice, speech translation across more than 70 languages, and Guided Vision audio descriptions. Google says camera visuals are not saved to Gemini History or used for model training or ads, bystander faces are redacted before Gemini processes them, and recording stops if the indicator light is covered. Source
5. Google Cloud’s Organization Policy MCP Server Reaches GA, With Gemini Authoring of Custom Constraints
Google Cloud announced general availability of a Model Context Protocol server for Organization Policy Service, along with natural-language authoring of custom constraints through Gemini. The MCP server exposes tools such as ListConstraints, GetCustomConstraint, GetEffectivePolicy, CreatePolicy and CreateCustomConstraint, so agents can audit posture, detect drift and manage exemptions across the resource hierarchy. Gemini writes the constraint from a prompt instead of requiring hand-written CEL, and can generate compliant and non-compliant test configurations for dry-run validation. Custom constraints cover Agent Runtime, Agent Search and Agent Studio on the Gemini platform as well as Cloud Run and GKE, with examples such as requiring agent identity or an approved Agent Gateway for deployed agents. Source