跳到正文
原文
The Verge · AI· Stevie Bonifield·· 3 天前AI 评分58

Google 发布 Gemini 4 Hydrogen 模型,称更擅长理解复杂问题

Google’s new Gemini model is better at understanding complicated questions

文章要点
  • 1Google 发布新模型 Gemini 4 Hydrogen,称其更擅长处理多部分问题、难懂错别字和高特异性请求。
  • 2模型今日起登陆搜索、Gemini 应用、AI Studio、Gemini API 等渠道,Google Search 产品副总裁称其已是 AI Mode 新默认模型。
  • 3官方称新模型相比 Gemini Flash 3.8 速度与智能显著跃升,达到或超过 Claude Sonnet 5.5、GPT-6.1 Terra 等同级模型,暂无独立验证。
  • 4思考档位新增 xhigh 选项,回答加入更多图片、对比表格、图表和地图,语言理解覆盖所有支持的 AI Mode 语言。
决策要点

实质是 Google 把新模型直接推上搜索默认入口,并将复杂问题理解作为主打能力,官方性能说法暂无独立验证。

  • 影响:企业可经 Gemini API、AI Studio 和 Gemini for business 直接用上新模型,复杂多部分问答类应用有换代选项;搜索侧 AI Mode 默认模型升级且回答更可视化,依赖搜索获取客户的业务需留意呈现形态变化。
  • 参考:可考虑在多步骤、输入易出错的实际场景用 Gemini API 做小规模对比实测,验证官方所称的复杂问题理解能力;值得盘点现有 Gemini 调用与思考档位配置,评估是否切换及选哪一档。
  • 依据:模型能力、性能对比与默认模型调整均来自 Google 及其 Search 产品副总裁的官方口径,暂无独立评测验证;可用范围与订阅限制以官方发布为准。
正文

Stevie Bonifield

Stevie Bonifield

is a news writer covering all things consumer tech. Stevie started out at Laptop Mag writing news and reviews on hardware, gaming, and AI.

Google says its new Gemini 4 Hydrogen model can give more helpful answers to complex questions, including “questions with multiple parts, hard-to-understand typos, or highly specific requests.” The new model can break down your questions and address each part in its responses. It’s available starting today in search, the Gemini app, all tiers of Google AI Studio, the Gemini API, Gemini for business, and in Google Antigravity for Google AI Pro and Ultra subscribers. Robby Stein, VP of product at Google Search, says it’s “now the new default model in AI Mode & being used for your hardest questions in AI Overviews.”

Google says Gemini 4 Hydrogen also gives users a wider range of links in its responses by “conducting simultaneous queries across all the relevant subtopics.” Its responses have more visuals, too, like images, comparison tables, graphs, and maps. One of Stein’s examples shows the AI generating an interactive animated driving simulation in response to a user’s question about how phones locate which lane of a highway a car is on in areas without cell service. Google says the new model also has improved language understanding “across all supported AI Mode languages.”

Users have more thinking level options with Gemini 4 Hydrogen, as well, including low, medium, high, and a new “xhigh” mode. According to Google, the new model “delivers significant leaps in both speed and intelligence” compared to Gemini Flash 3.8, and it “matches or outperforms some of the competing models in its class,” like Claude Sonnet 5.5 and GPT-6.1 Terra.

Follow topics and authors from this story to see more like this in your personalized homepage feed and to receive email updates.

  • Stevie Bonifield

来源:The Verge · AI · theverge.com