Google launches Gemini 4 Hydrogen as its lower-cost model
Reported by tradersunion, letsdatascience, Technobezz and 10 more outlets13 sources ↓
Latest in this story: Google brings Gemini 4 Hydrogen to Search worldwide
Google launched Gemini 4 Hydrogen for its Gemini app, Search and developer tools. Paid API pricing is $1.50 per million input tokens and $7.50 per million output tokens; Google says it scored 51 on Artificial Analysis’ index, versus 50 for Argon.
Key points
- Paid API use costs $1.50 per million input tokens and $7.50 per million output tokens
- Google reports an Artificial Analysis score of 51 for Hydrogen, compared with 50 for Argon
- Hydrogen is available in the Gemini app, Search, AI Studio and the Gemini API
- AndroidPure reports Hydrogen is now the default in Search AI Mode worldwide
Sources
- tradersunionRead ↗Google rolls out Gemini 4 Hydrogen across products and search
- letsdatascienceRead ↗Google rolls out Gemini 4 Hydrogen to replace 3.8 Flash
- TechnobezzRead ↗Google Launches Gemini 4 Hydrogen Across Gemini App, Search and API
- OfficeChaiRead ↗Google Releases Gemini 4 Hydrogen, Beats Sonnet 5.5 On Most Benchmarks
- Unite.AIRead ↗Google Launches Gemini 4 Hydrogen for Everyday Agentic Workflows
Details
Google describes Hydrogen as a faster, lower-cost model for coding and agentic tasks. It is available through the Gemini app, Google Search, AI Studio and the Gemini API; paid Google AI Pro and Ultra subscribers can also use it in Antigravity. SiliconANGLE reports that it is available to free Gemini app users as well.
Google says Hydrogen completed tasks 2.5 times faster than Gemini 3.8 Flash at the default setting. Its reported Artificial Analysis Intelligence Index score is 51 at the default high thinking level, compared with 50 for Gemini 4 Argon; the score rises to 54 at the new xhigh level. These figures are reported by Google, not presented as independent testing by the articles.
AndroidPure reports that Hydrogen is the default model in Search AI Mode worldwide and handles the hardest questions that trigger AI Overviews. Google says it is better at multipart and typo-filled queries and can surface more helpful links, but the article notes that these claims have not been independently tested.
Google also introduced Orchestrator in its Antigravity coding tool to switch between Hydrogen and Gemini 4 Carbon during a task. Google’s reported tests found results close to Carbon’s while using 24% less quota; Unite.AI reports near-Carbon task success on Terminal-Bench 4.1 and an average completion time 19% faster.
Written from coverage as of
How this story developed
- Google lets staff test a fast new Gemini model called Hydrogen
- Google rolls out Gemini 4 Hydrogen across its productsthis story
- Google brings Gemini 4 Hydrogen to Search worldwide
Images
- SiliconANGLE
- Unite.AI
- The Next Web
- AndroidPure
Stocks in this story
Change since open: 20 Nov to 21 Nov 2026
Last refresh