最新资讯
siliconangle.com OpenAI 据称以 1.25 万亿美元估值完成 110 亿美元新融资 dev.to Salesforce 推出 Koa Decisions,实现毫秒级 Agent 路由 bloomberg.com Neocloud 厂商 Groq 拟在 AI 抛售潮中将估值翻倍至 80 亿美元 techcrunch.com OpenAI 据称以 1.25 万亿美元估值追加融资 110 亿美元 developer.nvidia.com 在 NVIDIA Blackwell GPU 上加速 FLUX 3 Image Klein 图像生成 unite.ai OpenAI 推出 BioBasicsBench,测试 AI Agent 解读湿实验数据能力 www.youtube.com Warp 工程师展示让 AI Agent 自我复盘并修复配置的方法 theinformation.com Anthropic 据称下周启动 IPO 路演 unite.ai 科技股抛售后 Anthropic IPO 估值预期下调 dev.to AI 合同审查:依赖条款摘要前的 16 项失败测试 theguardian.com 美国起诉三人:涉嫌用 AI 冒充 CEO 诈骗逾 1 亿美元 x.com Anthropic 介绍 Claude 记忆机制:用“dreaming”整理经验
多元动态
登录 注册
大模型 www.aibase.com ·

谷歌调整 Gemini 订阅策略:10 月 9 日起免费版仅保留 Flash-Lite,付费层级进一步拉开

原文 我的收藏

据 AIbase 转述,Google 更新支持页面,宣布自 10 月 9 日起调整个人账户的 Gemini 模型访问与使用限制:免费用户仅能使用 Gemini 3.5-Flash-Lite,无法调用 Gemini 3.6 Flash 及更新模型;4.99 美元的 AI Plus 订阅用户仅限 Flash 和 Flash-Lite 轻量模型,不再支持 Gemini 3.1 Pro 等高级推理模型;19.99 美元起的 Google AI Pro 和 Ultra 订阅不受影响,仍可使用 Deep Think 和高级模型。此外 Google 引入低、中、高三档 Effort 工作负载模式,按任务计算复杂度实时扣减配额,高负载模式消耗更快。背景信息称 Google 近期在封闭环境测试 Gemini 4 "Argon" 模型,并持续迭代 Gemini 3.8 Flash 等轻量模型。

AI推荐理由 内容涉及 Google 对 Gemini 个人账户订阅与模型访问权限的调整,属于 AI 产品商业化与算力配额策略变化,对开发者和用户有实际参考价值。但来源为 AIbase 转述,正文未给出 Google 支持页面的直接链接或原文引用,部分模型版本号(如 Gemini 3.5-Flash-Lite、3.6 Flash、3.1 Pro、4 Argon)无法从输入中独立核实,且背景信息中的 Gemini 4 测试属于未确认传闻,需保留不确定性。

62

正文

Google Adjusts Gemini Subscription Strategy: Free Version Will Only Retain Flash-Lite Starting October 9, Paid Tiers Further Distinguish AIbase基地 Published inAI News · 3 min read · Oct 8, 2026 10 Google has recently updated its support page, announcing that starting October 9, it will adjust the access and usage restrictions for the Gemini model on personal accounts, tightening the model configuration for lower-tier users.. According to the new regulations, users who do not pay will only be able to access the Gemini 3.5-Flash-Lite version and will no longer have the ability to call the Gemini 3.6 Flash and newer models.;Users with a $4.99 AI Plus subscription will have their usage limited to the Flash and Flash-Lite lightweight models, and will no longer support advanced reasoning models such as Gemini 3.1 Pro.. This adjustment does not affect Google AI Pro and Ultra subscribers, who start at $19.99 per month, as they still enjoy the Deep Think option and the right to use advanced models.. In addition to downgrading model classifications, Google has also introduced three levels of workload (Effort) modes: low, medium, and high. The system will deduct usage in real-time based on the computational complexity of the task, with higher workload modes consuming quotas more quickly.。Background information shows that Google has recently started testing the new Gemini 4 "Argon" model in a closed environment and continues to frequently iterate on lightweight models such as Gemini 3.8 Flash.。 With the increasing costs of cutting-edge model inference and high concurrency computing loads, leading AI companies are gradually ending the "high-spec free trial" phase.Google's recent tiered and detailed quota control shifts high-computing models toward high-value paying users, marking a new stage in commercial large model applications, moving from "user scale expansion" to "fine-tuned operation of computing power economic benefits.". GeminiGoogleAI New WordsBrand Product Terms This article is from AIbase Daily Scan to view Welcome to the [AI Daily] column! This is your daily guide to exploring the world of artificial intelligence. Every day, we present you with hot topics in the AI field, focusing on developers, helping you understand technical trends, and learning about innovative AI product applications. —— Created by the AIbase Daily Team © Copyright AIbase Base 2024, Click to View Source -

来源报道 · 1 篇 · 以下链接均前往原始媒体

← 返回精选动态 原文

更多内容