AI News

Google 发布 Gemini 4 Hydrogen,称速度与智能均有大幅提升

9to5Google更新于 11月13日 03:26

Google announces Gemini 4 Hydrogen with ‘significant leaps’ in speed, intelligence

AI 导读

Google 发布 Gemini 4 Hydrogen,取代 Flash 品牌,称其为相对 Gemini 3.8 Flash 的独立大版本升级,支持原生 100 万 token 输入和最多 100 万 token 输出。官方称其在高默认设置下完成任务快 2.5 倍,Terminal-Bench 4.1 得分 57%(3.8 Flash 为 18.2%),编造性陈述减少 5 倍以上。该模型已进入 Gemini 应用模型选择器、Google 搜索 AI Mode 默认模型和 Antigravity,开发者调用价格为每百万输入 token 1.50 美元、每百万输出 token 7.50 美元。

推荐理由

精选AI 评分 90

原文列出 Hydrogen 相对 3.8 Flash 的速度、基准与智能体可靠性数据,可据此判断它在 Gemini 产品线中的位置。

主题

标签

正文

After this week’s Carbon availability, Google today unveiled Gemini 4 Hydrogen instead of keeping the “Flash” branding. This model is billed as bringing “cutting-edge intelligence to everyone.”

Benchmarks

Compared to Gemini 3.8 Flash (which launched in September), Hydrogen is billed as a “major standalone upgrade” with “significant leaps in both speed and intelligence.” It offers native 1 million-token input and up to 1 million output tokens. Specifically:

Hydrogen finishes “tasks 2.5 times faster at its high default setting”

“…more than [triples] Terminal-Bench 4.1 scores (57% vs. 18.2%)”

Cuts “fabricated claims by more than 5x”

Adapts “cleanly when a tool call fails or a search returns empty”

In terms of benchmarks:

“Hydrogen surpasses Argon on the Artificial Analysis Intelligence Index at its default high thinking level (51 vs. 50, and reaching 54 at its highest xhigh thinking setting), while often completing tasks faster than models in its class.”

“…our partners have seen Hydrogen run 2x to 2.8x faster while reducing token usage by 12% to 23%, compared to 3.8 Flash.”

“On multi-step tasks like deep research, data analysis, and broad codebase navigation, Hydrogen ranks #5 on Agent Arena, arena.ai’s live benchmark.”

“For deep domain reasoning and analysis, it leads its class on HLE-Diamond (58.8%), ranks #1 overall on Vals Tax Agent Bench (52.3%), ties for #1 on the Vals Legal Research Bench (55.8%), and leads its class on AutomationBench-AA.”

“It outperforms other models on the new Multimodal Knowledge Agent (MKA) benchmark (53.4%) for end-to-end agentic knowledge work over complex visual files, and on Agents’ Last Exam v1.1 (71.6%) for end-to-end professional workflows.”

Compared to 3.8 Flash, Hydrogen “reduces overreach in agentic tasks by up to 81% and under-service by up to 79%, and cuts misaligned actions and task shortcutting across tool-use and coding evaluations by 12x. When faced with impossible or blocked tasks, it is up to 2x less likely to bypass permission boundaries and achieves a 96.7% honesty rate on our internal agent honesty evaluation.”

Gemini app

In the model picker, 4 Hydrogen is billed as providing “Everyday help.”

Quick answers: 3.5 Flash-Lite

Everyday help: 4 Hydrogen

Detailed research: 3.1 Pro (on AI Pro)

Code & analysis: 4 Argon

Most complex work: 4 Carbon

Free users in the Gemini app get access to Hydrogen via the “[Auto] model selection when it’s the appropriate model for the prompt.”

Google Search

Google says Hydrogen significantly upgrades Search (AI Mode) by better understanding your intent:

…you can ask your question in your natural flow of thought — even if you have multiple parts to your query, hard-to-understand typos, or highly specific requests — and Search will intelligently comprehend what you’re looking for.

Responses get more visual and will feature a wider range of relevant links. Hydrogen is the default model for AI Mode and is used for “your hardest questions” in AI Overviews.

Antigravity

In Google Antigravity, 4 Hydrogen has an “Extra” thinking tier alongside Low, Medium, and High. Then there’s Orchestrator: “Carbon that may offload routine steps to Hydrogen. Same Carbon-level quality, fewer tokens used.”

To help make this multi-model collaboration automatic, we’re introducing Orchestrator in Google Antigravity. Rather than using a single model from start to finish, Orchestrator dynamically switches between models like Carbon and Hydrogen within the same session.

For workflows that require parallel agents, Carbon can act as the coordinator while Hydrogen can power worker and verifier agents to build and test changes simultaneously across isolated workspaces — helping teams ship complex projects faster without driving up token costs.

For developers, Hydrogen costs $1.50 per million input tokens and $7.50 per million output tokens.

Updating…

来源

9to5Google