PlatformsThe briefing
Google releases Gemini 4 Hydrogen model with broad availability
AI-generated illustration
PlatformsThe briefing
AI-generated illustration
Google launched Gemini 4 Hydrogen as the new default in AI Mode and AI Overviews. The model supports complex multi-part questions, wider link ranges, visuals, and four thinking levels including a new xhigh option. It is live today across Search, the Gemini app, AI Studio, the Gemini API, Gemini for business, and Google Antigravity for Pro and Ultra subscribers.
Google says Gemini 4 Hydrogen is now the default model in AI Mode and is used for the hardest questions in AI Overviews. It handles questions with multiple parts, hard-to-understand typos, or highly specific requests by breaking them down and addressing each part. The model also runs simultaneous queries across subtopics to return a wider range of links and adds more visuals such as images, comparison tables, graphs, and maps.
Access started today in Google Search, the Gemini app, all tiers of Google AI Studio, the Gemini API, Gemini for business, and Google Antigravity for Google AI Pro and Ultra subscribers. Google reports improved language understanding across all supported AI Mode languages and added a new xhigh thinking level alongside low, medium, and high options.
Google says the model delivers significant leaps in speed and intelligence compared with Gemini Flash 3.8 and matches or outperforms some competing models in its class. SiliconANGLE reports that on the Artificial Analysis Intelligence Index at default high thinking level, Hydrogen scored 51 while the frontier Gemini 4 Argon scored 50.
Simon Willison notes that Hydrogen is the new free tier model and powers AI features in Google Search. Pricing is set at $1.50 per million input tokens, $0.075 cached, and $7.50 output, which is a 25 percent discount to Carbon and Argon today and will become a 75 percent discount after December 1 when those models double in price.
Builders evaluating Gemini 4 Hydrogen should test it first on everyday code-writing and test-verification tasks where Google positions it as the high-efficiency execution engine. An illustrative scenario is a team that already uses Carbon or Argon for architectural decisions and routes routine implementation work to Hydrogen via the Gemini API; the 25 percent current discount and later 75 percent discount after December 1 would reduce spend if token usage stays similar. The evidence does not establish whether Hydrogen maintains the same token efficiency as Argon on complex code tasks, so teams must measure actual token spend and output quality on their own workloads before committing production traffic.
Operators reviewing Gemini 4 Hydrogen should confirm eligibility for their Google AI Pro or Ultra plan and review data-handling terms for Gemini for business before routing internal queries. An illustrative scenario is a company that wants to replace an existing paid model tier with the new free tier model in Search and the Gemini app; the change could lower direct model costs but requires checking whether the wider link and visual outputs meet internal compliance rules for customer-facing responses. The supplied evidence does not establish retention periods or exclusion-list handling for business-tier usage, so those terms must be verified directly with Google before any data migration.
Run a controlled side-by-side test of Gemini 4 Hydrogen against the current default model on a sample of 50 real user questions that include multi-part or typo-heavy inputs. Record answer completeness, number of links returned, presence of visuals, and any differences in thinking-level behavior. Compare the results against the documented rollout scope to decide whether to enable the new model for a pilot group of users or keep the prior default.
Source material behind this briefing. Our practical implications are Ashlr’s interpretation.
This briefing was prepared with AI assistance from the linked sources. The analysis may include interpretation; consult the original sources for complete context. Cover artwork is illustrative.

PlatformsThe briefing
AI-generated illustration
Google launches Tool Retrieval in the Gemini Interactions API. Agents can defer full tool schemas until needed, cutting token use while preserving selection accuracy for libraries of dozens or hundreds of tools.

PlatformsThe briefing
AI-generated illustration
Gemini for healthcare enters preview to connect EHRs, billing, guidelines and Medicare data for multi-step workflows under guardrails and human oversight.

PlatformsThe briefing
AI-generated illustration
Gemini 4 Carbon reaches all paying developers and enterprises today. A swarm feature on Antigravity lets one prompt spawn coordinated agents for long-horizon tasks.