Google has launched Gemini 4 Hydrogen, a fast new AI model built for everyday coding and agentic work, with a global rollout that started on 12 November 2026. According to Google, Hydrogen brings Gemini 4 intelligence to everyone, with speed and efficiency aimed at the tasks developers and businesses run most often.
Hydrogen is the execution-focused member of the Gemini 4 family. Google positions it for rapid work such as writing code, running tests, verifying changes, research and tool-using agents, while its sibling Carbon handles deeper reasoning. The launch follows Carbon’s general availability on 10 November and Argon’s on 13 October, completing a three-model lineup in under a month.
The model is rolling out across Google Search, the Gemini app, Google AI Studio, the Gemini API and Gemini for business. Free Gemini app users will receive Hydrogen through automatic model selection when Google’s router judges it the right fit for a prompt, while paid Google AI Pro and Ultra subscribers can also try it in Google Antigravity.
Gemini 4 Hydrogen at a Glance
| Detail | Gemini 4 Hydrogen |
|---|---|
| Announced | 12 November 2026 |
| Built for | Fast coding and agentic workflows |
| Based on | Gemini 4 Carbon, adapted for execution speed |
| Context window | 1,048,576 tokens, shared by input, thinking and output |
| Thinking levels | Low, medium, high (default) and new xhigh |
| API pricing | $1.50 per million input tokens, $7.50 per million output tokens |
| Free access | Rate-limited free tier on the Gemini API and AI Studio |
| Available in | Gemini API, AI Studio, Gemini app, Google Search, Antigravity, Gemini for business |
| Knowledge cutoff | March 2026 |
What Gemini 4 Hydrogen Is Built to Do
Hydrogen is designed for the work that fills a developer’s day rather than for showpiece reasoning tests. Google describes it as built for rapid execution across diverse agentic workflows, from writing and checking code to research tasks that call external tools. It accepts text, images, audio and video as input and produces text output.
The model sits alongside two siblings with different jobs. Argon, generally available since 13 October, is the flagship reasoning model, while Carbon, released on 10 November, supplies deeper reasoning inside Antigravity, Google’s software-development platform. Hydrogen takes the execution role, and Antigravity’s new Orchestrator switches between Hydrogen and Carbon during a task, using each where it fits best. Orchestrator is rolling out first to Ultra subscribers.
Google CEO Sundar Pichai posted about the launch on X, presenting Hydrogen as the model built for speed and efficiency across everyday tasks, coding and agentic workflows. The message is consistent across Google’s own announcement: this is a workhorse release, not a research showcase.
Gemini 4 Hydrogen Speed and Benchmarks
Speed is the headline claim. Google says Hydrogen completes tasks 2.5 times faster than its predecessor, Gemini 3.8 Flash, the model it now replaces as Google’s everyday coding and agent model. That comparison matters because Flash had become the default choice for high-volume, low-cost work.
On the Artificial Analysis Intelligence Index, Hydrogen scores 51 at its default high thinking level, against 50 for Argon, and reaches 54 at its highest xhigh thinking setting. The figures come from Google’s own announcement, and independent replication has not yet been published.
The picture is more mixed on coding benchmarks. On the cheating-adjusted Terminal-Bench 4.1 table in the model card, Hydrogen scores 57.0%, behind Claude Sonnet 5.5 at 63.9%. Google also notes that Hydrogen offers four thinking levels — low, medium, high by default and the new xhigh — so users can trade depth of reasoning against speed and cost for each task.
For context, the wider AI race keeps moving at a similar pace. This week also brought reports that GTA VI pre-orders pass $1 billion ahead of launch, a reminder that consumer technology demand remains strong even as the model wars dominate developer attention.
Gemini 4 Hydrogen Pricing and the 1 Million Token Window
Hydrogen carries one flat price across its full context window: $1.50 per million input tokens and $7.50 per million output tokens, with thinking tokens included in the output price. Batch and Flex processing cost half that, and cached input costs $0.075 per million tokens through implicit caching.
The context window holds 1,048,576 tokens shared by input, thinking and output, with 65,536 output tokens per request by default and more available through long decode. Google lists no per-minute limits for the model, though the free tier is rate-limited.
The pricing also sets a timetable for existing Flash users. Gemini 3.8 Flash rates remain $0.75 per million input tokens and $3.75 per million output tokens through 31 December 2026, then rise to Hydrogen’s rates on 1 January 2027. Developers still on Flash therefore have about six weeks to move their workloads across.
Where Gemini 4 Hydrogen Is Rolling Out
Google says Hydrogen is rolling out globally starting from the launch day, 12 November 2026. The rollout covers the Gemini API in AI Studio and Android Studio, the Gemini app, Google Search for all users, Gemini for business, and paid Google AI Pro and Ultra subscribers in Google Antigravity. Eligible Google Sheets subscriptions also get access through Gemini in Sheets.
Access still varies by surface. Free Gemini app users cannot pin Hydrogen in the model drop-down; they receive it only when automatic selection chooses it. In Antigravity, Hydrogen is rolling out to Pro and Ultra subscribers, while Orchestrator starts with Ultra only. The enterprise API model is a short-term generally available release, subject to at least 45 days’ notice before retirement, according to WinBuzzer’s coverage of the launch.
The Search rollout is the widest change for ordinary users. Hydrogen becomes the default model in AI Mode and handles harder questions in AI Overviews, according to The Verge’s coverage, which means many people will use the model without ever choosing it.
The Safety Picture Google Has Shared So Far
Google’s frontier-safety assessment found no material general-capability increase over Argon. Targeted testing placed Hydrogen below the critical capability levels for biological-information and cybersecurity risks, while reaching the alert thresholds in both domains. Google says the full Safety & Security Report will be available in the coming days.
That framing is worth reading carefully. Reaching an alert threshold is not the same as crossing a critical level, but it shows the model sits close enough to those boundaries for Google to flag it. The company has not published the full report yet, so independent scrutiny of the safety claims will have to wait.
What the Launch Means for the AI Race
Hydrogen’s arrival tightens the contest for the everyday developer workload. A flat $1.50/$7.50 price, a 1 million-token window and a 2.5x speed claim against Flash give Google a clear pitch to teams running agents at scale, while the Terminal-Bench gap behind Claude Sonnet 5.5 shows the coding crown is still contested.
The launch also underlines how quickly Google is shipping the Gemini 4 family. Three models reached general availability within a month, and the Orchestrator concept — switching models mid-task — points to a future where users pick outcomes rather than models. Whether that approach wins developers over will depend on real-world reliability, not launch-day benchmarks.
For now, the most concrete change is the simplest one. From this week, the model answering harder questions in Google Search and powering coding assistants in AI Studio is Hydrogen, and the SpaceX Starship Flight 15 catch shows the same week has delivered rapid progress on the hardware side too. The AI race, like the space race, is being run in public, one launch at a time.
Frequently Asked Questions
What is Gemini 4 Hydrogen?
Gemini 4 Hydrogen is Google’s fast execution model in the Gemini 4 family, launched on 12 November 2026 for everyday coding and agentic workflows. It is adapted from Gemini 4 Carbon and replaces Gemini 3.8 Flash as Google’s everyday coding and agent model.
How fast is Gemini 4 Hydrogen?
Google says Hydrogen completes tasks 2.5 times faster than Gemini 3.8 Flash. On the Artificial Analysis Intelligence Index it scores 51 at its default high thinking level and 54 at its highest xhigh setting, against 50 for Argon.
How much does Gemini 4 Hydrogen cost?
Paid API access costs $1.50 per million input tokens and $7.50 per million output tokens, flat across the full 1 million-token context window. Batch and Flex processing cost half that, and a rate-limited free tier is available on the Gemini API and in AI Studio.
Where can I use Gemini 4 Hydrogen?
Hydrogen is rolling out globally in the Gemini API, Google AI Studio, Android Studio, the Gemini app, Google Search, Gemini for business, Google Antigravity for Pro and Ultra subscribers, and Gemini in Sheets for eligible subscriptions. Free Gemini app users receive it through automatic model selection.
Does Gemini 4 Hydrogen replace Gemini 3.8 Flash?
Yes, as Google’s everyday coding and agent model. Flash pricing of $0.75 per million input tokens and $3.75 per million output tokens stays in place through 31 December 2026, then rises to Hydrogen’s rates on 1 January 2027.
