TUESDAY, NOVEMBER 17, 2026 Tech news, minus the noise.
N News Cast
LATEST
Home  /  Dev
Dev

GitHub Copilot Gemini 4 Hydrogen Joins the Model Menu

GitHub put Gemini 4 Hydrogen into every paid Copilot tier on November 13, the third frontier model to join the coding assistant in a single week.

GitHub Copilot Gemini 4 Hydrogen Joins the Model Menu
Source: Martin Vorel, CC BY-SA 4.0, via Wikimedia Commons

GitHub Copilot Gemini support keeps expanding

GitHub Copilot Gemini support took another step on November 13, when GitHub added Gemini 4 Hydrogen to every paid Copilot tier from Pro upward. It was the third frontier model to join the coding assistant in a single week, arriving days after Gemini 4 Carbon and Claude Fable 5.2. The model menu inside coding tools is growing fast, and with it the number of decisions a developer has to make before writing a line of code.

Hydrogen is Google's latest Flash-class model. It went generally available on November 12, carries a 1 million token context window, and is priced at $1.50 per million input tokens and $7.50 per million output tokens on the Gemini API. It is available on both free and paid API tiers, which makes it one of the cheapest frontier-class models you can call.

For Copilot users, the practical effect is simple: another strong, fast, cheap model is now one click away inside the tool they already use. That matters more than any benchmark, because the best model for a coding task is usually the one that is already open in your editor.

What Hydrogen is actually good at

Flash-class models are the workhorses of AI coding. They handle completions, quick edits, boilerplate, and routine refactors at a speed and price that heavier models cannot match. Hydrogen fits that role, with the large context window letting it hold a whole file or a small codebase in view while it works.

The contrast with its siblings is the point. Gemini 4 Carbon, which went generally available on November 10, is the heavier model for harder reasoning, and its API price is set to rise from $2/$10 to $4/$20 per million tokens after November 30. Claude Fable 5.2, released November 9, is Anthropic's flagship for long agentic coding sessions. Hydrogen is the fast, cheap option you reach for when the task does not need the flagship.

That three-way split is the story of this month in developer tools. Google's own Gemini 4 Argon release opened the family to everyone, and the Copilot integration means developers can now compare all three approaches without leaving their editor.

Admins get new controls too

The model expansion is not the only Copilot news this week. On November 16, GitHub announced that enterprise administrators can require selected Copilot customizations to come from plugins or managed settings instead of standalone user or project configuration. In plain terms, companies can now control which custom instructions and tools their developers' Copilot instances use, instead of letting every developer configure their own.

The same day, GitHub announced that Visual Studio Code 1.146 deprecates the Local harness, the older software that coordinates the model, context, and tools for Copilot agent sessions. Our earlier coverage of VS Code as an AI agent control center explains why that coordination layer matters: the harness is what turns a chat window into an agent that can run tools.

Both changes point the same way. As coding assistants get more powerful and more models join them, the organizations paying for them want more control over how they are configured. Choice for developers, guardrails for admins. That tension is now a product feature.

How to pick a model without overthinking it

The model menu is getting crowded, so here is a simple rule. Use Hydrogen, or any fast model, for completions, small edits, and anything you will review line by line anyway. Use Carbon or Fable 5.2 for the hard stuff: unfamiliar codebases, multi-file refactors, and long agent sessions where the model needs to hold a plan in its head. If you are not sure which one a task needs, start cheap and escalate only when the cheap model fails.

Also worth checking: the admin controls, if your team has them. A shared set of Copilot customizations beats ten developers each inventing their own, and the new plugin restrictions make that easier to enforce. The full Claude Fable 5.2 release notes are worth reading too, since Fable is the model most teams will compare Hydrogen against. The menu will keep growing. The skill is knowing when to stop choosing and start coding.

AS
News Cast Editorial Team

We cover AI, gadgets, startups and security — original reporting and explainers, written by humans, for humans. Got a tip? Reach us via the contact page.