Agent Upstart

Cool things people are building with agents — and practical help getting the most out of them.

Playbooks · Coding tasks

Gemini 4 Hydrogen is GA; there is no 4 Pro

Written November 13, 2026

Gemini 4 Hydrogen shipped November 12 at $1.50/$7.50. And the Pro label ends: 2.5 Pro requests become 3.8 Flash answers on December 1.

What shipped

Google introduced Gemini 4 Hydrogen on November 12, and it is available across Google's surfaces from that day: all tiers of Google AI Studio and the Gemini API, Gemini for business, the Gemini app, and Google Search, with Google Antigravity for paid Google AI Pro and Ultra subscribers. Google's announcement positions Hydrogen as the execution model in the Gemini 4 family — built for rapid work across agentic workflows and everyday coding — while Gemini 4 Carbon and Gemini 4 Argon remain the frontier models for deep, long-horizon reasoning. Google calls Hydrogen a standalone upgrade for teams migrating from Gemini 3.8 Flash.

Price and shape

The API model ID is gemini-4-hydrogen. Google's pricing page lists the standard paid rate at $1.50 per million input tokens and $7.50 per million output tokens, flat across the whole context window, with a free tier at no charge. The rest of the rate card:

On scores, Google reports that Hydrogen passes Argon on the Artificial Analysis Intelligence Index at its default high setting, 51 to 50, and reaches 54 at xhigh. Those are Google's published figures in its own announcement, carried here as the vendor's claim.

There is no Gemini 4 Pro

Anyone hunting the rate card for a Pro tier can stop. The Gemini 4 sections of Google's pricing page are exactly three: Carbon, Argon, and Hydrogen. No gemini-4-pro identifier appears on the pricing page, and the announcement names the same three models. The Pro label belonged to the older generations, and in the Gemini 4 family it does not exist. Code or notes that plan around a future 4 Pro are planning around a name Google has not published anywhere in its model or pricing documentation.

Built to pair with the frontier models

Google also pitches Hydrogen as the worker in a multi-model setup. Its announcement describes an Orchestrator in Google Antigravity that switches between models such as Carbon and Hydrogen inside one session — Carbon acting as coordinator, Hydrogen powering worker and verifier agents that build and test changes in parallel across isolated workspaces. On Terminal-Bench 4.0, Google says Orchestrator reached near-Carbon task success while consuming 24% less quota on average and finishing 19% faster. Those are Google's reported figures for its own harness, and they describe the intended division of labour: judgement from the expensive model, volume from the cheap one.

One more date on the same pricing page affects the comparison. Gemini 3.8 Flash, the model Hydrogen replaces, costs $0.75/$3.75 only through December 31, 2026; from January 1, 2027 it rises to $1.50/$7.50 — Hydrogen's exact price. Any saving from staying on 3.8 Flash is a six-week saving, not a steady state.

The December 1 change underneath

The quieter deadline rides along with this launch. From December 1, requests sent to gemini-2.5-pro are answered by gemini-3.8-flash instead: a different model, at a different price, returning different output, under the old name in your code. The migration detail lives in Gemini 2.5 Pro and Flash retire December 1. If a production system still names 2.5 Pro, that page is the one to read before the month turns — Hydrogen is the upgrade Google is pointing Flash-class workloads at, and the reroute will not ask first.

Related playbooks

Sources

Back to: Coding playbooks hub · Also: Gemini 2.5 Pro and Flash retire December 1 · Gemini CLI 0.69.0-preview.0: /stats heatmap for token use