Google Launches Gemini 4 Hydrogen, Replacing the Gemini 3.8 Flash Tier
The fast, lower-cost Gemini 4 tier arrived Nov 12, 2026 with flat API pricing of $1.50 per million input tokens and $7.50 per million output tokens, a 1 million token context window, and Google-reported speed figures that were not independently replicated at publication.
Key points
- Google released Gemini 4 Hydrogen on Thursday, Nov 12, 2026.
- Hydrogen replaces the Gemini 3.8 Flash tier across Google products.
- API pricing is flat at $1.50 per million input tokens and $7.50 per million output tokens across the entire context window.
Google released Gemini 4 Hydrogen on Thursday, Nov 12, 2026. Hydrogen is the fast, lower-cost tier of the Gemini 4 family. It sits below Gemini 4 Carbon, the strongest tier, and Gemini 4 Argon.
Hydrogen replaces the Gemini 3.8 Flash tier across Google products. It now powers Google Search globally and the Gemini app. It is also available in the Gemini API, Google AI Studio, Android Studio and Antigravity for AI Pro and Ultra subscribers, according to reports of the launch in Unite.AI and Silicon Report.
Flat pricing across the full context window
API pricing is flat at $1.50 per million input tokens and $7.50 per million output tokens across the entire context window. Cached input costs $0.075 per million tokens through implicit caching. Artificial Analysis confirmed the pricing independently, Silicon Report reported.
The model carries a 1 million token unified context window, 1,048,576 tokens in full. It offers four thinking levels: low, medium, high and a new xhigh level. High is the default.
Flat pricing across the whole window is the detail developers will read first. The price does not step up as a conversation or a document set grows toward the full million tokens, and the cached input rate of $0.075 per million sets the floor for repeated work on the same material.
Speed and benchmark claims, attributed to Google
Google says Hydrogen completes tasks 2.5 times faster than Gemini 3.8 Flash at its default high thinking level. That figure is a Google claim. The company benchmark figures in the announcement are vendor-reported and were not independently replicated at publication.
On Google's cheating-adjusted Terminal-Bench 4.1 table in the model card, Hydrogen scores 57.0 percent. That sits below Claude Sonnet 5.5 at 63.9 percent on the same table.
One independent score is available. Artificial Analysis gave Hydrogen at its xhigh level a composite 54 on its Intelligence Index v5.0.1.
A short-term release with a March 2026 cutoff
Google Cloud's model page describes Hydrogen as a short-term generally available release. It is subject to a minimum 45-day retirement notice. Its knowledge cutoff is March 2026.
Readers weighing a switch from the Flash tier get three hard facts from the launch itself: the flat $1.50 and $7.50 token prices, the 1,048,576 token window, and the replacement of Gemini 3.8 Flash inside Search and the Gemini app. The speed figure and the benchmark table are Google's own numbers, and the one independent index score published so far is the Artificial Analysis composite of 54.
Timeline
- Mar 2026Knowledge cutoff for Gemini 4 Hydrogen, according to Google Cloud's model page.
- Nov 12, 2026Google released Gemini 4 Hydrogen, the fast, lower-cost tier of the Gemini 4 family, at flat API pricing of $1.50 per million input tokens and $7.50 per million output tokens.
- Nov 12, 2026Hydrogen replaced the Gemini 3.8 Flash tier across Google products, including Google Search globally and the Gemini app.