Model releases · · 2 min read
Gemini 4 Hydrogen arrives in Google Search and replaces Flash
Google has put Gemini 4 Hydrogen into Search and retired the Flash line. Here is what shipped, how it benchmarks, and what growth teams should check.
Takeaway Gemini 4 Hydrogen now powers Google Search, so re-check how your brand and content appear in AI answers and test cost per task before switching tools.
What shipped
Google has rolled out Gemini 4 Hydrogen and brought it into Google Search. The announcement shows example cards from the Search interface. Ars Technica reports that the model replaces 3.8 Flash, and The Decoder describes it as the successor to the whole Flash series. Nine outlets covered the release.
What is new
The notable change is the pace. Ars Technica says Hydrogen is not getting the slow, security-conscious rollout that Carbon and Argon received. One coverage headline also points to changes in coding, Search and the Gemini app. The announcement summary does not give a detailed feature list; see the source.
Benchmarks (developer-reported)
The figures below come from The Decoder, citing the independent Artificial Analysis Intelligence Index rather than Google itself.
- Hydrogen scores higher than the larger Argon model, only four weeks after Argon launched.
- It comes close to Claude Sonnet 5.5 while costing less.
- Per task, it costs roughly two and a half times as much as GPT-6.1 Sol.
- It still falls short on agentic coding.
Price and access
The source does not list prices. It confirms that the model is available in Google Search, and coverage also mentions the Gemini app. Check Google's documentation for developer pricing and limits.
What it means for marketing
First, treat Search as a changed surface. With gemini 4 hydrogen behind it, AI-generated answers may summarize and cite your pages differently than last month. Re-run your priority queries, note whether your brand is mentioned, and compare against competitors.
Second, review content that depends on being quoted. Clear, factual pages with specific answers are easier for a model to use accurately. Log any shifts in impressions and clicks so you can separate model effects from seasonal noise.
Third, be careful with tool choices. Hydrogen looks strong on general scores, but the higher per-task cost against GPT-6.1 Sol and the weaker agentic coding result matter if your team builds automations or scripts. Run a small test on your own workflows, such as briefs, ad variants or reporting, and measure cost per finished task before moving budget.
Finally, expect more changes. A faster rollout means less lead time, so assign someone to monitor Search behavior weekly.
Source
Google: Google Search gets a major upgrade with Gemini 4 Hydrogen (2026-11-12)
Also reported by: The Decoder, Ars Technica
Written by the AI CMO OS newsroom from the publisher's report. Facts and figures are as reported by the source; read the original for full details.
More in Model releases
Seedance 2.5 ByteDance video AI now has a 4K API for developers
ByteDance's Volcano Engine has opened API access to the 4K Seedance 2.5 video model. Here is what the release means for marketing and growth teams.
Huawei Cloud plans open source for CloudRobo-WM1.2 world action model
Huawei Cloud released the CloudRobo-WM1.2 embodied world action model and says it will go open source. Here is what that means for growth teams.
Alibaba flagship model Qwen4-Max cuts task costs 60%, adds cyber guardrails
Alibaba launched Qwen4-Max, its most advanced model, claiming 60% lower task costs and stricter cyber safeguards. What growth teams should check before testing it.