Gemini 4 Hydrogen is 2.5x faster than 3.8 Flash, and costs twice as much for now

• Everyone: in the Gemini app's model menu and in Google Search. Free Gemini app users get it through auto model selection when Gemini decides it fits the prompt.
• Developers: all tiers of the Gemini API and Google AI Studio, plus the paid tier in Android Studio.
• Paid Google AI Pro and Ultra subscribers: Google Antigravity and Gemini in Google Sheets.
• Businesses: Gemini for business, Antigravity, and Gemini in Sheets for eligible plans.

What Google says it does. All of these numbers are Google's:

• Finishes tasks 2.5x faster than Gemini 3.8 Flash at its default "high" thinking setting.
• More than triples 3.8 Flash's Terminal-Bench 4.1 score (57% vs. 18.2%) and cuts fabricated claims by more than 5x.
• Scores 51 on the Artificial Analysis Intelligence Index at its default setting, one point above Argon's 50, and 54 at its top "xhigh" setting.
• Four thinking levels (low, medium, high and the new xhigh), so you can trade answer quality for speed and cost.
• Takes up to 1 million tokens in and can write up to 1 million tokens out.
• In early partner tests on SQL- and data-heavy work, Google says it ran 2x to 2.8x faster than 3.8 Flash and used 12% to 23% fewer tokens.

Google is also adding Orchestrator to Antigravity. It switches between Carbon and Hydrogen within one session. On Terminal-Bench 4.0, Google says Orchestrator came close to Carbon's success rate while using 24% less quota and finishing 19% faster.

The price. The Gemini API pricing page (opens in a new tab) lists Hydrogen at $1.50 per million input tokens and $7.50 per million output tokens on the paid tier. Thinking tokens count as output. Google says that price is flat across the full 1 million-token window. Batch and Flex cost half: $0.75 in and $3.75 out. A free tier exists in the API, and on the free tier Google uses your content to improve its products.

The catch. Gemini 3.8 Flash costs $0.75 in and $3.75 out through Dec. 31, so Hydrogen costs twice as much per token for now. On Jan. 1, 2027, 3.8 Flash goes up to $1.50 and $7.50, the same as Hydrogen. Hydrogen also supports only implicit caching; explicit caching isn't available. And every speed and benchmark figure above comes from Google. We haven't tested it.

Our take. If you already run agents on 3.8 Flash, Hydrogen is worth a test today. Measure total cost per finished task, not cost per token: a faster model that needs fewer retries and fewer tokens can still come out ahead. If your workload is cheap, high-volume calls, the Flash promo price is hard to beat until Dec. 31.

Not tested by MODEL. Not sponsored. -- MODEL Editors