Google released Gemini 4 Hydrogen on November 12, its fast model in place of the Flash line. It is the default in AI Mode, sits in the Gemini app's model menu, and costs $1.50 and $7.50 per million tokens on the paid API, double Gemini 3.8 Flash's price until January 1. Who should switch and who should wait.
Google’s new fast model is live everywhere at once. Gemini 4 Hydrogen, released November 12, takes the slot the Flash models held. It now answers in Google Search, sits in the Gemini app’s model menu, and is open to developers on the Gemini API at $1.50 per million input tokens and $7.50 per million output tokens. If you use Gemini for everyday writing and research, you get it today without changing plans. If you pay for the API, it costs twice what Gemini 3.8 Flash costs until the end of the year.
What Google released
Hydrogen is Google’s answer for speed. The heavier Gemini 4 models, Argon and Carbon, stay the picks for long, hard reasoning jobs. Google drops the Flash name with this model, a change 9to5Google pointed out.
Where it shows up, according to Google’s announcement:
- Google Search. Hydrogen is rolling out to all users globally. Google’s Search team calls it the biggest model upgrade to Search yet. 9to5Google reports it is the default model for AI Mode and handles the hardest questions in AI Overviews.
- Gemini app. Google says everyone can start using it from the model menu. 9to5Google reports that free users get it when the Auto setting picks it for a prompt.
- Developers. Free and paid tiers of Google AI Studio and the Gemini API, plus the paid tier in Android Studio.
- Work accounts. Gemini for business, through the API or the model menu, and Gemini in Google Sheets on eligible business plans.
- Antigravity. Google’s coding tool gets Hydrogen for paid AI Pro and Ultra subscribers, plus a new Orchestrator mode that hands routine steps from Carbon to Hydrogen inside one session.
What it costs
Google’s Gemini API pricing page lists Hydrogen’s standard paid tier at $1.50 per million input tokens and $7.50 per million output tokens, flat across the full 1 million token context window. Thinking tokens bill as output. Cached input costs $0.075 per million, and Batch and Flex requests cost half, at $0.75 and $3.75. The free tier covers it too, with Google allowed to use free-tier data to improve its products.
The comparison that matters is Gemini 3.8 Flash. The same page lists it at $0.75 and $3.75 through December 31, 2026, then $1.50 and $7.50 from January 1, 2027. So Hydrogen launches at the price Flash moves to in seven weeks.
In consumer plans, nothing new costs money. Hydrogen comes with the free app and with the plans you already pay for. Separately, from November 13 the $4.99 AI Plus plan loses the Pro models, as our AI Plus report explains.
What Google claims it does better
Every number here is Google’s own. Independent tests are not out yet.
- Speed. Google says Hydrogen finishes tasks 2.5 times faster than 3.8 Flash at its default setting, and that partners saw it run 2 to 2.8 times faster with 12% to 23% fewer tokens on data-heavy work.
- Coding. Its Terminal-Bench 4.1 score more than triples, from 18.2% to 57%, according to Google.
- Accuracy. Google reports more than 5 times fewer fabricated claims than 3.8 Flash.
- Ranking. At its default high setting it scores 51 on the Artificial Analysis Intelligence Index, one point above Argon, and 54 at its new xhigh setting.
- Settings. Four thinking levels: low, medium, high (the default) and xhigh. You pay for the depth you pick.
What to do this week
- If you use Gemini or AI Mode for email drafts, research and summaries, do nothing. The upgrade reaches you on its own. Re-run one task that frustrated you before and see if it holds up better.
- If your app runs on Gemini 3.8 Flash, test Hydrogen now and keep paying Flash prices until December. Run your own prompts on both. If Hydrogen’s speed and token savings hold on your workload, plan the switch before January 1, when both cost the same.
- If you run Batch jobs, check the half-price tier. At $0.75 and $3.75, Batch Hydrogen matches today’s standard Flash price.
- If your company uses Google Workspace, ask your admin whether Hydrogen is on in Sheets for your plan. Google says eligible business subscriptions only.
Flash users pay double now or the same price in January
This section is our inference. Google has made its fast model better and, on paper, more expensive. Read together, the two price lines on Google’s page say the real increase was already scheduled. Gemini 3.8 Flash doubles on January 1 whether you move or not. That turns Hydrogen into a seven-week decision about timing for most API customers, as long as Google’s speed claims hold on real work.
The token claims matter more than the sticker price. Output tokens, including thinking, drive most bills. A model that finishes in fewer tokens can cost less per task at a higher per-token rate. That is the test to run before December 31, and the Google Cloud spend caps are worth turning on while you run it.