Google 推出 Gemini 4 Hydrogen，取代 3.8 Flash
Google rolls out Gemini 4 Hydrogen, finally replacing 3.8 Flash
Google 发布 Gemini 4 Hydrogen，并已在自家产品中取代 Gemini 3.8 Flash，访问限制比此前 Argon、Carbon 更宽松，发布即可使用
同一新闻，精选展示《Google DeepMind 发布 Gemini 4 Hydrogen 模型》
同一件事另外 6 家也报道了
- Google DeepMind··首发
- Hacker News · 300+ 热帖··+1 分钟
- IT之家 · AI··+16 分钟
谷歌发布 Gemini 4 Hydrogen 模型，主打日常编程与智能体工作流
新增信息：输入每百万Token 1.50美元、输出每百万Token 7.50美元的成本。
- Hacker News · 300+ 热帖··+51 分钟
Google 在 Antigravity 中推出 Gemini 4 Hydrogen 与 Orchestrator
新增信息：与Orchestrator模型选项结合，并与Gemini 4 Carbon集成。
- Ars Technica · AI··+1 小时 22 分·当前阅读
Google 推出 Gemini 4 Hydrogen，取代 3.8 Flash
新增信息：取代了Gemini 3.8 Flash，并且访问限制比Argon和Carbon更宽松。
- TechCrunch · AI··+1 小时 27 分
Google 推出 Gemini 4 Hydrogen 模型提升搜索复杂问题理解能力
新增信息：提升了理解复杂搜索问题的能力，速度比前代快2.5倍且成本更低。
- The Decoder··+1 小时 58 分
- MarkTechPost··+15 小时 45 分
按发布时间排列 · 事件后续出现时自动追加 · 由模型根据各家标题与摘要比对
Google is continuing its elemental AI naming scheme with the new Gemini 4 Hydrogen. These names are less descriptive than the Pro/Flash monikers of earlier releases, but Hydrogen's purpose is clear: It replaces the faster, more economical Flash models most people in Google's ecosystem use.
With Google's recent Argon and Carbon releases, the rollout was staged and began with cybersecurity researchers. Even when general availability began, Google restricted access to users paying for its most expensive AI subscription. The limits for Hydrogen are much more relaxed, with access beginning immediately. This model has replaced Gemini 3.8 Flash across the company's products.
Google claims that Hydrogen represents a major improvement versus the previous Flash model. It runs in one of four thinking levels (low, medium, high, and the new xhigh), and even in high mode, it's 2.5 times faster than Flash 3.8. Plus, it reduces "fabricated claims" by more than 5x, according to Google's analysis. Users will also see gains in code generation, with Hydrogen more than tripling 3.8 Flash's score in Terminal-Bench. In some cases, Hydrogen even trades blows with Argon. It managed a score of 54 in the popular Artificial Analysis Intelligence Index with xhigh mode versus 50 for Argon.
来源: Ars Technica · AI · arstechnica.com