Google has introduced Gemini 4 Hydrogen, a new model that the company says brings cutting-edge intelligence to everyday coding and agentic workflows with a focus on speed and efficiency. Announced shortly after the availability of Carbon, Hydrogen replaces the "Flash" branding used for earlier lightweight models, and Google positions it alongside its frontier models, Gemini 4 Carbon and Gemini 4 Argon, which are built for deep, long-horizon reasoning.
Google says Hydrogen surpasses Argon on the Artificial Analysis Intelligence Index at its default high thinking level, scoring 51 against Argon's 50 and reaching 54 at its highest xhigh setting, while often completing tasks faster than models in its class. The model supports a native 1 million-token input window and up to 1 million output tokens, and is priced flat across that context at $1.50 per million input tokens and $7.50 per million output tokens.S1S2
Compared with Gemini 3.8 Flash, which launched on September 2, Google describes Hydrogen as a major standalone upgrade. It finishes tasks 2.5 times faster at its high default setting, more than triples Terminal-Bench 4.1 scores (57% against 18.2%), cuts fabricated claims by more than five times, and recovers cleanly when a tool call fails or a search returns empty. Partners reported running Hydrogen two to 2.8 times faster while reducing token usage by 12% to 23%.S1S2
On specialised tests, Google reports that Hydrogen ranks fifth on Agent Arena, arena.ai's live benchmark for multi-step tasks such as deep research and codebase navigation. It leads its class on HLE-Diamond with 58.8%, ranks first overall on the Vals Tax Agent Bench with 52.3%, and ties for first on the Vals Legal Research Bench with 55.8%. It also outperforms other models on the new Multimodal Knowledge Agent benchmark at 53.4% and on Agents' Last Exam v1.1 at 71.6%. Google adds that, relative to 3.8 Flash, the model reduces overreach in agentic tasks by up to 81% and under-service by up to 79%, and achieves a 96.7% honesty rate on an internal agent honesty evaluation.S1S2
In the Gemini app, Hydrogen is positioned for everyday help, with better adherence to tone, format and rules and clearer tables and visual breakdowns, so that email drafts, study guides and rewrites come back ready to use. The model picker places it between quick answers on 3.5 Flash-Lite and detailed research on 3.1 Pro, and free users will reach it through automatic model selection when it is the right fit for a prompt. For multi-model workflows, Google is also introducing an Orchestrator in Google Antigravity that switches between Carbon and Hydrogen within a session; on Terminal-Bench 4.0 it reached near-Carbon task success while consuming 24% less quota and completing 19% faster.S1S2
