News in short

• Google has launched Gemini 4 Hydrogen, a faster AI model built for everyday coding and agentic workflows.
• The model is rolling out globally across Google Search, the Gemini app, AI Studio, Gemini API and business products.
• Gemini 4 Hydrogen supports four thinking levels: low, medium, high and xhigh.
• Google lists API pricing at $1.50 per million input tokens and $7.50 per million output tokens.
• Google Antigravity is also getting Orchestrator, which can switch between Gemini models during a task.

Google has introduced Gemini 4 Hydrogen, a new AI model designed for faster everyday work, coding and agentic workflows. The company says the model combines strong reasoning with quicker execution. It is also expanding Gemini 4 access across consumer, developer and enterprise products.

Gemini 4 Hydrogen is positioned below Google’s more advanced Gemini 4 Carbon and Gemini 4 Argon models for demanding, long-running reasoning tasks. Instead, Hydrogen focuses on rapid execution, software development, research and tasks that involve multiple tools.

Gemini 4 Hydrogen focuses on speed and everyday AI tasks

Google says Hydrogen can work independently or alongside Carbon and Argon in multi-model workflows. It supports repository coding, debugging, web development, data analysis and multimodal document work. The model can process PDFs, spreadsheets, charts, audio and recorded video within its large context window.

The company claims Hydrogen completes tasks 2.5 times faster than Gemini 3.8 Flash at its default high thinking level. Google also reports a Terminal-Bench 4.1 score of 57 percent, compared with 18.2 percent for Gemini 3.8 Flash. These are Google’s own benchmark results and may not reflect performance across every workload.

Hydrogen offers four thinking levels: low, medium, high and xhigh. Users and developers can choose the level that best balances response quality, speed and cost. Google says the model scored 51 on the Artificial Analysis Intelligence Index at its default high setting, rising to 54 at xhigh.

Google Antigravity gets Orchestrator for multi-model workflows

Google is also introducing Orchestrator in Google Antigravity. The feature can switch between models such as Gemini 4 Carbon and Hydrogen during the same session. This allows a more capable model to handle planning, while Hydrogen takes care of coding, testing and verification.

Google says Orchestrator achieved near-Carbon task success rates on Terminal-Bench 4.0. It reportedly used 24 percent less quota and completed tasks 19 percent faster on average. The company also says teams can run parallel worker and verifier agents across separate workspaces.

Gemini 4 Hydrogen availability and API pricing

Gemini 4 Hydrogen is rolling out globally across Google AI Studio, the Gemini API, Gemini for business, the Gemini app and Google Search. Paid Google AI Pro and Ultra subscribers can also access it in Google Antigravity. Eligible business users can use the model through supported Google products, including Gemini in Google Sheets.

In the Gemini API, Google has set pricing at $1.50 per million input tokens and $7.50 per million output tokens. The rates apply across the model’s one-million-token context window, according to the company.

For Gemini app users, access depends on the plan and model selection. Google says free users will receive Hydrogen through automatic model selection when it suits a prompt. Availability across individual features may vary by product and subscription.

Google says Hydrogen also improves safeguards for agentic tasks, including reducing unintended actions and making the model less likely to bypass permission boundaries. The launch broadens access to the Gemini 4 family, while giving developers a faster option for routine execution. For more context, read our report on Gemini 4 Carbon and its developer-focused capabilities.