{"@context":"https://schema.org","@type":"NewsArticle","generatedAt":"2026-07-23T06:40:50.084Z","headline":"Google 发布 Gemini 3.6 Flash 与 3.5 Flash-Lite，降低企业 AI 智能体成本","description":"Google 发布 Gemini 3.6 Flash 与 3.5 Flash-Lite，旨在降低企业 AI 智能体的延迟与 token 成本。3.6 Flash 输出 token 比前代减少 17%，在 DeepSWE 上成功率达 49%（前代 37%），定价 $1.50/1M 输入 token。","url":"https://www.aioga.com/news/cmruv44fg008zbii0eui6tcvw/","mainEntityOfPage":"https://www.aioga.com/news/cmruv44fg008zbii0eui6tcvw/","datePublished":"2026-07-21T16:06:55.000Z","dateModified":"2026-07-21T16:06:55.000Z","inLanguage":"zh-CN","publisher":{"@type":"NewsMediaOrganization","name":"Aioga","url":"https://www.aioga.com"},"citation":["https://www.artificialintelligence-news.com/news/googles-gemini-3-6-flash-targets-enterprise-agent-token-costs","https://aihot.virxact.com/items/cmruv44fg008zbii0eui6tcvw"],"canonicalUrl":"https://www.aioga.com/news/cmruv44fg008zbii0eui6tcvw/","directAnswer":{"@type":"Answer","text":"Aioga 编辑摘要：Google 发布 Gemini 3.6 Flash 与 3.5 Flash-Lite，旨在降低企业 AI 智能体的延迟与 token 成本。 Aioga 将其归入「模型更新」方向，重点关注它对真实使用和行业竞争的影响。","url":"https://www.aioga.com/news/cmruv44fg008zbii0eui6tcvw/","dateCreated":"2026-07-21T16:06:55.000Z","author":{"@type":"Organization","@id":"https://www.aioga.com/authors/aioga-editorial/#editorial-team","name":"Aioga Editorial Team","url":"https://www.aioga.com/authors/aioga-editorial/"}},"evidence":[{"@type":"CreativeWork","name":"artificialintelligence-news.com source article","url":"https://www.artificialintelligence-news.com/news/googles-gemini-3-6-flash-targets-enterprise-agent-token-costs","datePublished":"2026-07-21T16:06:55.000Z","provider":{"@type":"Organization","name":"artificialintelligence-news.com","url":"https://www.artificialintelligence-news.com/news/googles-gemini-3-6-flash-targets-enterprise-agent-token-costs"}},{"@type":"CreativeWork","name":"AIHot archive record","url":"https://aihot.virxact.com/items/cmruv44fg008zbii0eui6tcvw","datePublished":"2026-07-21T16:06:55.000Z","provider":{"@type":"Organization","name":"AIHot","url":"https://aihot.virxact.com/items/cmruv44fg008zbii0eui6tcvw"}}],"aggregationSource":"Artificial Intelligence News（RSS）","originalPublisher":{"name":"artificialintelligence-news.com","url":"https://www.artificialintelligence-news.com/news/googles-gemini-3-6-flash-targets-enterprise-agent-token-costs"},"article":{"id":"cmruv44fg008zbii0eui6tcvw","slug":"cmruv44fg008zbii0eui6tcvw","url":"https://www.aioga.com/news/cmruv44fg008zbii0eui6tcvw/","title":"Google 发布 Gemini 3.6 Flash 与 3.5 Flash-Lite，降低企业 AI 智能体成本","title_en":"Google's Gemini 3.6 Flash targets enterprise agent token costs","summary":"Google 发布 Gemini 3.6 Flash 与 3.5 Flash-Lite，旨在降低企业 AI 智能体的延迟与 token 成本。3.6 Flash 输出 token 比前代减少 17%，在 DeepSWE 上成功率达 49%（前代 37%），定价 $1.50/1M 输入 token。","source":"Artificial Intelligence News（RSS）","sourceUrl":"https://www.artificialintelligence-news.com/news/googles-gemini-3-6-flash-targets-enterprise-agent-token-costs","aiHotUrl":"https://aihot.virxact.com/items/cmruv44fg008zbii0eui6tcvw","publishedAt":"2026-07-21T16:06:55.000Z","category":"模型更新","score":65,"selected":false,"articleBody":["Google：https://blog.google/ has released Gemini 3.6 Flash and 3.5 Flash-Lite as new workhorses designed to cut latency and token costs for enterprise AI agents.","The economics of running autonomous software agents inside a production environment come down to a fixed equation few vendors advertise directly. A model needs to reason through a multi-step task competently, but every extra token it generates while doing so adds cost and delay to a workflow that might run thousands of times an hour.","Teams building background agents rather than chat interfaces need throughput first and parameter count second. Google’s answer, announced this week, splits that trade-off across three models: Gemini 3.6 Flash for coding and multimodal reasoning, Gemini 3.5 Flash-Lite for high-volume, low-latency work, and a restricted Gemini 3.5 Flash Cyber variant built for vulnerability remediation.","Google’s developer documentation for 3.6 Flash centres on one figure: 17 percent fewer output tokens than the prior 3.5 Flash version, based on measurements from the Artificial Analysis Index.","In specific synthetic tests, including the Datacurve DeepSWE benchmark, Google reports drops in token usage of up to 65 percent. Pricing sits at $1.50/1M input tokens and $7.50/1M output tokens, positioning the model for reasoning loops that run continuously rather than on-demand.","On DeepSWE, the company records a 49 percent success rate for 3.6 Flash against 37 percent for its predecessor. On MLE Bench, the score moves from 49.7 percent to 63.9 percent, and on Google’s GDPval-AA v2 test – which attempts to measure real-world knowledge work rather than coding puzzles – 3.6 Flash scores 1421 against 1349 for the older model.","Figma has integrated 3.6 Flash into its prototyping infrastructure, and according to Matt Colyer, the company’s Director of Product Engineering, the model gives developers a faster route through design iterations without a drop in output quality.","Legal technology platform Harvey and research tool Hebbia route data through the model for multimodal document work: ingesting raw financial filings, parsing document structure, reading embedded charts, and producing draft reports for review.","Google also folded a client-side computer-use tool directly into the Gemini API and Gemini Enterprise platforms, removing the custom intermediary software engineers previously built to let models operate on top of an operating system.","The company reports an OSWorld-Verified score of 83.0 percent, up from 78.4 percent, and says updated safeguards against chemical, biological, radiological, and nuclear misuse improve resistance to jailbreaking without raising refusal rates for benign requests.","Gemini 3.5 Flash-Lite targets a different job: document processing and agentic search running at volume rather than reasoning depth. The Artificial Analysis Index measured the model at 350 output tokens per second, the fastest in the 3.5 series according to Google.","Pricing runs at $0.3/1M input tokens and $2.5/1M output tokens, cheap enough that engineering teams can route simple, high-volume subagent requests to a minimal thinking level and reserve higher thinking levels for multi-step work.","On Google’s GDM-MRCR v2 long-context test, Gemini 3.5 Flash-Lite recorded a 72.2 percent success rate against 60.1 percent for its predecessor, and its GDPval-AA v2 score nearly doubled, from 642 to 1140. The model carries the same native computer-use tool as 3.6 Flash.","Separately, Google says Gemini 3.5 Pro remains in partner testing ahead of a full release, and pre-training for the next Gemini 4 architecture is already underway.","Automated vulnerability scanners now surface flaws faster than most security teams can patch them, and that gap is where Google positions Gemini 3.5 Flash Cyber.","The model is built to validate and remediate code vulnerabilities, and Google reports performance on the CyberGym benchmark competitive with frontier models (though it hasn’t made those figures public in the same detail as its consumer-facing releases.)","Distribution stays restricted to governments and vetted partners through a pilot programme, a limitation Google frames as a safeguard against the model generating exploit code for offensive use.","Inside Google’s CodeMender security agent, multiple instances of 3.5 Flash Cyber run in parallel, cross-checking one another’s findings before producing a single remediation report a human reviewer signs off on.","Engineering teams seeking to integrate these new models can access them through the Gemini API via Google AI Studio, Android Studio, or the Gemini Enterprise Agent Platform. Consumers can also access the new models in the Gemini app and 3.5 Flash-Lite is also rolling out in Google Search.","See also: Bristol Myers Squibb buys Nvidia AI system for drug discovery ：https://www.artificialintelligence-news.com/news/bristol-myers-squibb-nvidia-ai-system-drug-discovery/","Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo：https://www.ai-expo.net/?utm_source=AI-News&utm_medium=Footer-banner&utm_campaign=world-series taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx：https://techexevent.com/?utm_source=AI-News&utm_medium=Footer-banner&utm_campaign=world-series and is co-located with other leading technology events including the Cyber Security & Cloud Expo：https://cybersecuritycloudexpo.com/?utm_source=CloudTech-News&utm_medium=Footer-banner&utm_campaign=world-series. Click here：https://techexevent.com/?utm_source=AI-News&utm_medium=Footer-banner&utm_campaign=world-series for more information.","AI News is powered by TechForge Media：https://techforge.pub/?utm_source=AI-News&utm_medium=Footer-banner&utm_campaign=world-series. Explore other upcoming enterprise technology events and webinars here：https://techforge.pub/events/?utm_source=AI-News&utm_medium=Footer-banner&utm_campaign=world-series.","Ryan Daws：https://www.artificialintelligence-news.com/news/author/ryan/","Physical AI：https://www.artificialintelligence-news.com/categories/ai-and-us/physical-ai/","AI Business Strategy：https://www.artificialintelligence-news.com/categories/inside-ai/ai-business-strategy/, Artificial Intelligence：https://www.artificialintelligence-news.com/categories/artificial-intelligence/, Features：https://www.artificialintelligence-news.com/categories/features/, Finance AI：https://www.artificialintelligence-news.com/categories/ai-in-action/finance-ai/, World of Work：https://www.artificialintelligence-news.com/categories/ai-and-us/world-of-work/","Sponsored Content：https://www.artificialintelligence-news.com/categories/sponsored-content/","Healthcare & Wellness AI：https://www.artificialintelligence-news.com/categories/ai-in-action/healthcare-wellness-ai/","Artificial Intelligence：https://www.artificialintelligence-news.com/categories/artificial-intelligence/","AI in Action：https://www.artificialintelligence-news.com/categories/ai-in-action/","All our premium content and latest tech news delivered straight to your inbox","AI News is part of TechForge：https://techforge.pub/"],"articleImages":[{"sourceUrl":"https://secure.gravatar.com/avatar/d82963cecdd93f33733ed383f2d8005d91790f3740b49a926d4c180557eec721?s=96&d=mm&r=g","alt":"","afterParagraph":21,"url":"/media/articles/cmruv44fg008zbii0eui6tcvw/5ce3cf2821211c69.png"},{"sourceUrl":"https://www.artificialintelligence-news.com/wp-content/uploads/2026/06/Airia-webinar.png","alt":"","afterParagraph":25,"url":"/media/articles/cmruv44fg008zbii0eui6tcvw/09f94645ae1f29f3.png"},{"sourceUrl":"https://www.artificialintelligence-news.com/wp-content/uploads/2025/01/300x600-150x300.png","alt":"","afterParagraph":25,"url":"/media/articles/cmruv44fg008zbii0eui6tcvw/84efdfaa93523a46.png"},{"sourceUrl":"https://www.artificialintelligence-news.com/wp-content/uploads/2026/07/Bristol-Myers-Squibb-buys-Nvidia-AI-system-for-drug-discovery-scaled-e1784603783132.jpg","alt":"Bristol Myers Squibb buys Nvidia AI system for drug discovery","afterParagraph":25,"url":"/media/articles/cmruv44fg008zbii0eui6tcvw/4d06a9de75c52106.jpg"},{"sourceUrl":"https://www.artificialintelligence-news.com/wp-content/uploads/2026/07/1784472098908.jpeg","alt":"","afterParagraph":26,"url":"/media/articles/cmruv44fg008zbii0eui6tcvw/6d66e21b1adeb114.jpg"},{"sourceUrl":"https://www.artificialintelligence-news.com/wp-content/uploads/2026/07/US-public-health-agencies-to-test-OpenAI-and-Anthropic-AI-models-scaled-e1784536548513.jpg","alt":"US public health agencies to test OpenAI and Anthropic AI models","afterParagraph":27,"url":"/media/articles/cmruv44fg008zbii0eui6tcvw/985d55971985bdce.jpg"}],"mediaStatus":"ok","articleBodyZh":["Google：https://blog.google/ 已发布 Gemini 3.6 Flash 和 3.5 Flash-Lite，作为旨在降低企业 AI 代理延迟和令牌成本的新型主力模型。","在生产环境中运行自主软件代理的经济学归结为一个少数供应商会直接宣传的固定公式。模型需要能够胜任多步骤任务的推理，但在执行过程中生成的每一个额外令牌都会增加成本和延迟，而这些工作流可能每小时运行数千次。","构建后台代理而非聊天界面的团队首先需要吞吐量，其次才是参数数量。Google 本周宣布的解决方案，将这一权衡分配到三个模型：用于编码和多模态推理的 Gemini 3.6 Flash，用于高容量、低延迟工作的 Gemini 3.5 Flash-Lite，以及用于漏洞修复的受限 Gemini 3.5 Flash Cyber 版本。","Google 面向开发者的 3.6 Flash 文档集中展示了一个指标：相比之前的 3.5 Flash 版本，输出令牌减少了 17%，此数据来自人工分析指数（Artificial Analysis Index）的测量。","在特定的合成测试中，包括 Datacurve DeepSWE 基准测试，Google 报告令牌使用量最多下降了 65%。定价为每百万输入令牌 1.50 美元，每百万输出令牌 7.50 美元，将模型定位于持续运行的推理循环，而非按需运行。","在 DeepSWE 测试中，3.6 Flash 的成功率为 49%，其前身为 37%。在 MLE Bench 测试中，分数从 49.7% 提升至 63.9%，而在 Google 的 GDPval-AA v2 测试中——该测试尝试衡量现实世界知识工作而非编码难题——3.6 Flash 得分为 1421，旧型号为 1349。","Figma 已将 3.6 Flash 集成到其原型设计基础设施中，据该公司产品工程总监 Matt Colyer 称，该模型为开发者提供了更快的设计迭代路径，同时不会降低输出质量。","法律技术平台 Harvey 和研究工具 Hebbia 通过该模型处理多模态文档工作：摄取原始财务文件、解析文档结构、读取嵌入图表，并生成供审阅的报告草稿。","谷歌还将一个客户端计算机使用工具直接集成到 Gemini API 和 Gemini Enterprise 平台中，移除了以前为让模型在操作系统上运行而构建的自定义中间软件。","公司报告称，OSWorld-Verified 得分为 83.0%，高于之前的 78.4%，并表示针对化学、生物、放射性和核滥用的更新防护措施，提高了对越狱的抵抗能力，同时没有提高对无害请求的拒绝率。","Gemini 3.5 Flash-Lite 针对不同的任务：文档处理和大量运行的代理搜索，而不是深度推理。人工分析指数测量该模型的输出速度为每秒 350 个 token，根据谷歌，这是 3.5 系列中最快的。","定价为每 1M 输入 token $0.3，输出 token $2.5，价格足够低，使工程团队可以将简单、高量的子代理请求路由到最低思考级别，并为多步骤工作保留更高级的思考级别。","在谷歌的 GDM-MRCR v2 长上下文测试中，Gemini 3.5 Flash-Lite 的成功率为 72.2%，而其前代为 60.1%，其 GDPval-AA v2 得分几乎翻倍，从 642 增至 1140。该模型携带与 3.6 Flash 相同的原生计算机使用工具。","此外，谷歌表示 Gemini 3.5 Pro 仍在合作伙伴测试中，尚未全面发布，而针对下一代 Gemini 4 架构的预训练已在进行中。","自动化漏洞扫描器现在能比大多数安全团队更快地发现漏洞，而谷歌正将 Gemini 3.5 Flash Cyber 定位于填补这一差距。","该模型旨在验证和修复代码漏洞，谷歌报告其在 CyberGym 基准测试中的表现与前沿模型具有竞争力（尽管它没有像面向消费者的版本那样公开这些数据的详细信息）。","分发依然限制在政府和经过审查的合作伙伴通过试点计划获取，谷歌将这一限制描述为防止模型生成用于进攻用途的漏洞利用代码的一种安全措施。","在谷歌的 CodeMender 安全代理中，多实例的 3.5 Flash Cyber 并行运行，互相交叉核对发现项，然后生成由人工审核签署的单一修复报告。","希望整合这些新模型的工程团队可以通过 Google AI Studio、Android Studio 或 Gemini 企业代理平台访问 Gemini API 来使用它们。消费者也可以在 Gemini 应用中访问这些新模型，同时 3.5 Flash-Lite 也正在 Google 搜索中推出。","另请参见：Bristol Myers Squibb 购买 Nvidia AI 系统用于药物发现：https://www.artificialintelligence-news.com/news/bristol-myers-squibb-nvidia-ai-system-drug-discovery/","想了解更多来自行业领导者的 AI 和大数据资讯吗？查看 AI & Big Data Expo：https://www.ai-expo.net/?utm_source=AI-News&utm_medium=Footer-banner&utm_campaign=world-series，该展会将在阿姆斯特丹、加州和伦敦举行。该综合活动是 TechEx 的一部分：https://techexevent.com/?utm_source=AI-News&utm_medium=Footer-banner&utm_campaign=world-series，并与其他领先的技术活动共同举办，包括 Cyber Security & Cloud Expo：https://cybersecuritycloudexpo.com/?utm_source=CloudTech-News&utm_medium=Footer-banner&utm_campaign=world-series。点击这里获取更多信息：https://techexevent.com/?utm_source=AI-News&utm_medium=Footer-banner&utm_campaign=world-series","AI News 由 TechForge Media 提供支持：https://techforge.pub/?utm_source=AI-News&utm_medium=Footer-banner&utm_campaign=world-series。在此了解其他即将举行的企业技术活动和网络研讨会：https://techforge.pub/events/?utm_source=AI-News&utm_medium=Footer-banner&utm_campaign=world-series","Ryan Daws：https://www.artificialintelligence-news.com/news/author/ryan/","物理 AI：https://www.artificialintelligence-news.com/categories/ai-and-us/physical-ai/","AI 业务策略：https://www.artificialintelligence-news.com/categories/inside-ai/ai-business-strategy/，人工智能：https://www.artificialintelligence-news.com/categories/artificial-intelligence/，专题报道：https://www.artificialintelligence-news.com/categories/features/，金融 AI：https://www.artificialintelligence-news.com/categories/ai-in-action/finance-ai/，职场世界：https://www.artificialintelligence-news.com/categories/ai-and-us/world-of-work/","赞助内容：https://www.artificialintelligence-news.com/categories/sponsored-content/","医疗保健与健康 AI：https://www.artificialintelligence-news.com/categories/ai-in-action/healthcare-wellness-ai/","人工智能：https://www.artificialintelligence-news.com/categories/artificial-intelligence/","AI 实践：https://www.artificialintelligence-news.com/categories/ai-in-action/","我们所有的优质内容和最新科技新闻将直接发送到您的收件箱","AI 新闻是 TechForge 的一部分：https://techforge.pub/"],"translationStatus":"translated","bodyOrigin":"source-page","editorial":{"summary":"Aioga 编辑摘要：Google 发布 Gemini 3.6 Flash 与 3.5 Flash-Lite，旨在降低企业 AI 智能体的延迟与 token 成本。 Aioga 将其归入「模型更新」方向，重点关注它对真实使用和行业竞争的影响。","background":"背景分析：模型与研究类动态需要结合能力边界、开放方式、成本、可用性和真实任务表现判断，单项指标领先不等于已经形成稳定采用。","viewpoint":"Aioga 判断：这条动态更适合作为行业观察信号，当前信息足以建立线索，但不足以推导长期结论。","implications":"影响分析：对相关团队而言，短期应先核对来源、可用范围和实际成本，再判断是否值得接入或跟进。","nextStep":"后续观察：继续观察官方文档、实际可用性、价格变化、开发者反馈和竞品回应。","evidenceRefs":["title","summary","articleBody"],"confidence":"medium","status":"published","aiGenerated":false,"autoApproved":true,"generatedBy":"rule-safe-fallback","generatedAt":"2026-07-23T06:49:19.070Z","sourceHash":"6e18c5766fe4b989","validation":{"passed":true,"mode":"rule-safe-fallback","checks":["schema","length","source-attribution","no-html"]}},"tags":["模型更新","Artificial Intelligence News（RSS）"],"translations":{"zh-CN":{"title":"Google 发布 Gemini 3.6 Flash 与 3.5 Flash-Lite，降低企业 AI 智能体成本","summary":"Google 发布 Gemini 3.6 Flash 与 3.5 Flash-Lite，旨在降低企业 AI 智能体的延迟与 token 成本。3.6 Flash 输出 token 比前代减少 17%，在 DeepSWE 上成功率达 49%（前代 37%），定价 $1.50/1M 输入 token。","category":"模型更新","source":"artificialintelligence-news.com","aggregationSource":"Artificial Intelligence News（RSS）","pageTitle":"Google 发布 Gemini 3.6 Flash 与 3.5 Flash-Lite，降低企业 AI 智能体成本 - Aioga AI资讯","description":"Google 发布 Gemini 3.6 Flash 与 3.5 Flash-Lite，旨在降低企业 AI 智能体的延迟与 token 成本。3.6 Flash 输出 token 比前代减少 17%，在 DeepSWE 上成功率达 49%（前代 37%），定价 $1.50/1M 输入 token。","url":"https://www.aioga.com/news/cmruv44fg008zbii0eui6tcvw/"},"en":{"title":"Google releases Gemini 3.6 Flash and 3.5 Flash-Lite, reducing the cost of enterprise AI agents","summary":"Google released Gemini 3.6 Flash and 3.5 Flash-Lite, aimed at reducing latency and token costs for enterprise AI agents. Gemini 3.6 Flash outputs 17% fewer tokens than the previous generation, achieves a success rate of 49% on DeepSWE (compared to 37% for the previous generation), and is priced at $1.50 per 1M input tokens.","category":"Models","source":"Artificial Intelligence News（RSS）","aggregationSource":"Artificial Intelligence News（RSS）","pageTitle":"Google releases Gemini 3.6 Flash and 3.5 Flash-Lite, reducing the cost of enterprise AI agents - Aioga AI News","description":"Google released Gemini 3.6 Flash and 3.5 Flash-Lite, aimed at reducing latency and token costs for enterprise AI agents. Gemini 3.6 Flash outputs 17% fewer tokens than the previous...","url":"https://www.aioga.com/en/news/cmruv44fg008zbii0eui6tcvw/","contentTranslated":true,"sourceHash":"57d4fef2ae1d2858","translatedAt":"2026-07-22T19:24:06.792Z"},"ja":{"title":"Google、Gemini 3.6 Flashと3.5 Flash-Liteを発表、企業のAIエージェントのコストを削減","summary":"Google は Gemini 3.6 Flash と 3.5 Flash-Lite を発表し、企業向け AI エージェントの遅延とトークンコストを削減することを目指しています。3.6 Flash の出力トークンは前世代より17%減少し、DeepSWE での成功率は49%（前世代 37%）に達し、価格は入力トークン1百万あたり $1.50 です。","category":"モデル更新","source":"Artificial Intelligence News（RSS）","aggregationSource":"Artificial Intelligence News（RSS）","pageTitle":"Google、Gemini 3.6 Flashと3.5 Flash-Liteを発表、企業のAIエージェントのコストを削減 - Aioga AIニュース","description":"Google は Gemini 3.6 Flash と 3.5 Flash-Lite を発表し、企業向け AI エージェントの遅延とトークンコストを削減することを目指しています。3.6 Flash の出力トークンは前世代より17%減少し、DeepSWE での成功率は49%（前世代 37%）に達し、価格は入力トークン1百万あたり $1.50 です。","url":"https://www.aioga.com/ja/news/cmruv44fg008zbii0eui6tcvw/","contentTranslated":true,"sourceHash":"57d4fef2ae1d2858","translatedAt":"2026-07-22T19:24:05.865Z"},"ko":{"title":"구글, Gemini 3.6 Flash와 3.5 Flash-Lite 출시로 기업 AI 지능체 비용 절감","summary":"Google은 기업 AI 에이전트의 지연 시간과 토큰 비용을 줄이기 위해 Gemini 3.6 Flash 및 3.5 Flash-Lite를 출시했습니다. 3.6 Flash는 이전 세대보다 토큰 출력이 17% 감소했으며, DeepSWE에서 성공률은 49%(이전 세대 37%)에 달하며, 가격은 입력 토큰 1M당 $1.50입니다.","category":"모델 업데이트","source":"Artificial Intelligence News（RSS）","aggregationSource":"Artificial Intelligence News（RSS）","pageTitle":"구글, Gemini 3.6 Flash와 3.5 Flash-Lite 출시로 기업 AI 지능체 비용 절감 - Aioga AI 뉴스","description":"Google은 기업 AI 에이전트의 지연 시간과 토큰 비용을 줄이기 위해 Gemini 3.6 Flash 및 3.5 Flash-Lite를 출시했습니다. 3.6 Flash는 이전 세대보다 토큰 출력이 17% 감소했으며, DeepSWE에서 성공률은 49%(이전 세대 37%)에 달하며, 가격은 입력 토큰 1M당 $1.50입니다.","url":"https://www.aioga.com/ko/news/cmruv44fg008zbii0eui6tcvw/","contentTranslated":true,"sourceHash":"57d4fef2ae1d2858","translatedAt":"2026-07-22T19:24:58.826Z"},"es":{"title":"Google lanza Gemini 3.6 Flash y 3.5 Flash-Lite, reduciendo el costo de los agentes de IA para empresas","summary":"Google lanzó Gemini 3.6 Flash y 3.5 Flash-Lite, con el objetivo de reducir la latencia y el costo por token de los agentes de IA empresariales. Gemini 3.6 Flash produce un 17% menos de tokens que la generación anterior, logra una tasa de éxito del 49% en DeepSWE (frente al 37% de la generación anterior) y tiene un precio de $1.50 por 1 millón de tokens de entrada.","category":"Modelos","source":"Artificial Intelligence News（RSS）","aggregationSource":"Artificial Intelligence News（RSS）","pageTitle":"Google lanza Gemini 3.6 Flash y 3.5 Flash-Lite, reduciendo el costo de los agentes de IA para empresas - Aioga Noticias de IA","description":"Google lanzó Gemini 3.6 Flash y 3.5 Flash-Lite, con el objetivo de reducir la latencia y el costo por token de los agentes de IA empresariales. Gemini 3.6 Flash produce un 17% meno...","url":"https://www.aioga.com/es/news/cmruv44fg008zbii0eui6tcvw/","contentTranslated":true,"sourceHash":"57d4fef2ae1d2858","translatedAt":"2026-07-22T19:24:48.750Z"},"fr":{"title":"Google publie Gemini 3.6 Flash et 3.5 Flash-Lite, réduisant le coût des agents IA pour les entreprises","summary":"Google publie Gemini 3.6 Flash et 3.5 Flash-Lite, visant à réduire la latence et le coût en tokens des agents IA pour les entreprises. Gemini 3.6 Flash produit 17% de tokens en moins que la génération précédente, atteint un taux de réussite de 49% sur DeepSWE (contre 37% pour la génération précédente), avec un tarif de 1,50 $/1M de tokens d'entrée.","category":"Modèles","source":"Artificial Intelligence News（RSS）","aggregationSource":"Artificial Intelligence News（RSS）","pageTitle":"Google publie Gemini 3.6 Flash et 3.5 Flash-Lite, réduisant le coût des agents IA pour les entreprises - Aioga Actualités IA","description":"Google publie Gemini 3.6 Flash et 3.5 Flash-Lite, visant à réduire la latence et le coût en tokens des agents IA pour les entreprises. Gemini 3.6 Flash produit 17% de tokens en moi...","url":"https://www.aioga.com/fr/news/cmruv44fg008zbii0eui6tcvw/","contentTranslated":true,"sourceHash":"57d4fef2ae1d2858","translatedAt":"2026-07-22T19:25:39.917Z"},"de":{"title":"Google veröffentlicht Gemini 3.6 Flash und 3.5 Flash-Lite, senkt die Kosten für KI-Intelligenzagenten in Unternehmen","summary":"Google veröffentlicht Gemini 3.6 Flash und 3.5 Flash-Lite, um die Latenz und Token-Kosten von Unternehmens-KI-Agenten zu senken. Gemini 3.6 Flash reduziert die Ausgabe-Token im Vergleich zur Vorgängerversion um 17%, erreicht eine Erfolgsquote von 49 % bei DeepSWE (vorher 37 %) und wird mit 1,50 $/1 Mio. Eingabe-Token bepreist.","category":"模型更新","source":"Artificial Intelligence News（RSS）","aggregationSource":"Artificial Intelligence News（RSS）","pageTitle":"Google veröffentlicht Gemini 3.6 Flash und 3.5 Flash-Lite, senkt die Kosten für KI-Intelligenzagenten in Unternehmen - Aioga KI-News","description":"Google veröffentlicht Gemini 3.6 Flash und 3.5 Flash-Lite, um die Latenz und Token-Kosten von Unternehmens-KI-Agenten zu senken. Gemini 3.6 Flash reduziert die Ausgabe-Token im Ver...","url":"https://www.aioga.com/de/news/cmruv44fg008zbii0eui6tcvw/","contentTranslated":true,"sourceHash":"57d4fef2ae1d2858","translatedAt":"2026-07-22T19:25:48.670Z"},"pt-BR":{"title":"O Google lança Gemini 3.6 Flash e 3.5 Flash-Lite, reduzindo o custo de agentes de IA corporativos","summary":"O Google lançou o Gemini 3.6 Flash e o 3.5 Flash-Lite, com o objetivo de reduzir a latência e o custo de tokens dos agentes de IA corporativos. O Gemini 3.6 Flash gera 17% menos tokens em comparação com a versão anterior, alcançando uma taxa de sucesso de 49% no DeepSWE (ante 37% da versão anterior), com preço de $1,50 por 1 milhão de tokens de entrada.","category":"模型更新","source":"Artificial Intelligence News（RSS）","aggregationSource":"Artificial Intelligence News（RSS）","pageTitle":"O Google lança Gemini 3.6 Flash e 3.5 Flash-Lite, reduzindo o custo de agentes de IA corporativos - Aioga Notícias de IA","description":"O Google lançou o Gemini 3.6 Flash e o 3.5 Flash-Lite, com o objetivo de reduzir a latência e o custo de tokens dos agentes de IA corporativos. O Gemini 3.6 Flash gera 17% menos to...","url":"https://www.aioga.com/pt-BR/news/cmruv44fg008zbii0eui6tcvw/","contentTranslated":true,"sourceHash":"57d4fef2ae1d2858","translatedAt":"2026-07-22T19:26:31.944Z"},"ru":{"title":"Google выпустила Gemini 3.6 Flash и 3.5 Flash-Lite, снижая затраты на корпоративных ИИ-агентов","summary":"Google выпустила Gemini 3.6 Flash и 3.5 Flash-Lite, с целью снижения задержки и стоимости токенов для корпоративных ИИ-агентов. В Gemini 3.6 Flash количество выводимых токенов уменьшилось на 17% по сравнению с предыдущей версией, уровень успешности на DeepSWE достиг 49% (прежний показатель 37%), цена $1.50 за 1 млн входных токенов.","category":"模型更新","source":"Artificial Intelligence News（RSS）","aggregationSource":"Artificial Intelligence News（RSS）","pageTitle":"Google выпустила Gemini 3.6 Flash и 3.5 Flash-Lite, снижая затраты на корпоративных ИИ-агентов - Aioga Новости ИИ","description":"Google выпустила Gemini 3.6 Flash и 3.5 Flash-Lite, с целью снижения задержки и стоимости токенов для корпоративных ИИ-агентов. В Gemini 3.6 Flash количество выводимых токенов умен...","url":"https://www.aioga.com/ru/news/cmruv44fg008zbii0eui6tcvw/","contentTranslated":true,"sourceHash":"57d4fef2ae1d2858","translatedAt":"2026-07-22T19:26:30.582Z"},"ar":{"title":"أصدرت Google نظامي Gemini 3.6 Flash و3.5 Flash-Lite لتقليل تكلفة الذكاء الاصطناعي للمؤسسات","summary":"أصدرت Google الإصدارين Gemini 3.6 Flash و 3.5 Flash-Lite، بهدف تقليل زمن الاستجابة وتكاليف الرموز لوكلاء الذكاء الاصطناعي المؤسسيين. إنتاج الرموز في 3.6 Flash يقل بنسبة 17٪ مقارنة بالإصدار السابق، ونسبة النجاح على DeepSWE تصل إلى 49٪ (كانت 37٪ في الإصدار السابق)، وسعره 1.50 دولار لكل مليون رمز مدخل.","category":"模型更新","source":"Artificial Intelligence News（RSS）","aggregationSource":"Artificial Intelligence News（RSS）","pageTitle":"أصدرت Google نظامي Gemini 3.6 Flash و3.5 Flash-Lite لتقليل تكلفة الذكاء الاصطناعي للمؤسسات - Aioga أخبار الذكاء الاصطناعي","description":"أصدرت Google الإصدارين Gemini 3.6 Flash و 3.5 Flash-Lite، بهدف تقليل زمن الاستجابة وتكاليف الرموز لوكلاء الذكاء الاصطناعي المؤسسيين. إنتاج الرموز في 3.6 Flash يقل بنسبة 17٪ مقارنة...","url":"https://www.aioga.com/ar/news/cmruv44fg008zbii0eui6tcvw/","contentTranslated":true,"sourceHash":"57d4fef2ae1d2858","translatedAt":"2026-07-22T19:27:21.196Z"},"hi":{"title":"गूगल ने Gemini 3.6 फ्लैश और 3.5 फ्लैश-लाइट जारी किया, जिससे उद्यमों के एआई इंटेलिजेंट एजेंट की लागत कम हुई","summary":"गूगल ने Gemini 3.6 Flash और 3.5 Flash-Lite जारी किया, जिसका उद्देश्य उद्यम AI एजेंटों की देरी और टोकन लागत को कम करना है। 3.6 Flash पिछले संस्करण की तुलना में 17% कम टोकन आउटपुट करता है, DeepSWE पर सफल दर 49% है (पिछले संस्करण 37%), और मूल्य $1.50/1M इनपुट टोकन है।","category":"模型更新","source":"Artificial Intelligence News（RSS）","aggregationSource":"Artificial Intelligence News（RSS）","pageTitle":"गूगल ने Gemini 3.6 फ्लैश और 3.5 फ्लैश-लाइट जारी किया, जिससे उद्यमों के एआई इंटेलिजेंट एजेंट की लागत कम हुई - Aioga AI समाचार","description":"गूगल ने Gemini 3.6 Flash और 3.5 Flash-Lite जारी किया, जिसका उद्देश्य उद्यम AI एजेंटों की देरी और टोकन लागत को कम करना है। 3.6 Flash पिछले संस्करण की तुलना में 17% कम टोकन आउटपुट कर...","url":"https://www.aioga.com/hi/news/cmruv44fg008zbii0eui6tcvw/","contentTranslated":true,"sourceHash":"57d4fef2ae1d2858","translatedAt":"2026-07-22T19:27:21.842Z"},"it":{"title":"Google rilascia Gemini 3.6 Flash e 3.5 Flash-Lite, riducendo i costi degli agenti AI aziendali","summary":"Google ha rilasciato Gemini 3.6 Flash e 3.5 Flash-Lite, con l'obiettivo di ridurre la latenza e il costo dei token degli agenti AI aziendali. Gemini 3.6 Flash produce il 17% di token in meno rispetto alla generazione precedente, ha un tasso di successo del 49% su DeepSWE (contro il 37% della generazione precedente) e il prezzo è di $1,50/1M token in ingresso.","category":"模型更新","source":"Artificial Intelligence News（RSS）","aggregationSource":"Artificial Intelligence News（RSS）","pageTitle":"Google rilascia Gemini 3.6 Flash e 3.5 Flash-Lite, riducendo i costi degli agenti AI aziendali - Aioga Notizie IA","description":"Google ha rilasciato Gemini 3.6 Flash e 3.5 Flash-Lite, con l'obiettivo di ridurre la latenza e il costo dei token degli agenti AI aziendali. Gemini 3.6 Flash produce il 17% di tok...","url":"https://www.aioga.com/it/news/cmruv44fg008zbii0eui6tcvw/","contentTranslated":true,"sourceHash":"57d4fef2ae1d2858","translatedAt":"2026-07-22T19:28:03.636Z"},"nl":{"title":"Google lanceert Gemini 3.6 Flash en 3.5 Flash-Lite, verlaagt de kosten van AI-agents voor bedrijven","summary":"Google heeft Gemini 3.6 Flash en 3.5 Flash-Lite uitgebracht, gericht op het verlagen van de latentie en tokenkosten van AI-agents voor bedrijven. 3.6 Flash genereert 17% minder tokens dan de vorige generatie, behaalt een succespercentage van 49% op DeepSWE (vorige generatie 37%) en wordt geprijsd op $1,50 per 1 miljoen invoertokens.","category":"模型更新","source":"Artificial Intelligence News（RSS）","aggregationSource":"Artificial Intelligence News（RSS）","pageTitle":"Google lanceert Gemini 3.6 Flash en 3.5 Flash-Lite, verlaagt de kosten van AI-agents voor bedrijven - Aioga AI-nieuws","description":"Google heeft Gemini 3.6 Flash en 3.5 Flash-Lite uitgebracht, gericht op het verlagen van de latentie en tokenkosten van AI-agents voor bedrijven. 3.6 Flash genereert 17% minder tok...","url":"https://www.aioga.com/nl/news/cmruv44fg008zbii0eui6tcvw/","contentTranslated":true,"sourceHash":"57d4fef2ae1d2858","translatedAt":"2026-07-22T19:28:02.714Z"},"tr":{"title":"Google, Gemini 3.6 Flash ve 3.5 Flash-Lite'ı yayınladı, kurumsal yapay zeka ajanlarının maliyetini düşürdü","summary":"Google, işletme AI ajanlarının gecikmesini ve token maliyetlerini azaltmayı amaçlayarak Gemini 3.6 Flash ve 3.5 Flash-Lite'ı yayınladı. 3.6 Flash, önceki nesle göre çıktı token sayısını %17 oranında azalttı, DeepSWE'de başarı oranı %49'a ulaştı (önceki nesil %37) ve fiyatlandırması 1M giriş token için 1,50$ olarak belirlendi.","category":"模型更新","source":"Artificial Intelligence News（RSS）","aggregationSource":"Artificial Intelligence News（RSS）","pageTitle":"Google, Gemini 3.6 Flash ve 3.5 Flash-Lite'ı yayınladı, kurumsal yapay zeka ajanlarının maliyetini düşürdü - Aioga AI Haberleri","description":"Google, işletme AI ajanlarının gecikmesini ve token maliyetlerini azaltmayı amaçlayarak Gemini 3.6 Flash ve 3.5 Flash-Lite'ı yayınladı. 3.6 Flash, önceki nesle göre çıktı token say...","url":"https://www.aioga.com/tr/news/cmruv44fg008zbii0eui6tcvw/","contentTranslated":true,"sourceHash":"57d4fef2ae1d2858","translatedAt":"2026-07-22T19:28:52.099Z"},"vi":{"title":"Google ra mắt Gemini 3.6 Flash và 3.5 Flash-Lite, giảm chi phí cho các AI agent doanh nghiệp","summary":"Google ra mắt Gemini 3.6 Flash và 3.5 Flash-Lite, nhằm giảm độ trễ và chi phí token của các tác nhân AI doanh nghiệp. Gemini 3.6 Flash tạo ra token ít hơn 17% so với thế hệ trước, tỷ lệ thành công trên DeepSWE đạt 49% (thế hệ trước 37%), định giá $1,50/1 triệu token đầu vào.","category":"模型更新","source":"Artificial Intelligence News（RSS）","aggregationSource":"Artificial Intelligence News（RSS）","pageTitle":"Google ra mắt Gemini 3.6 Flash và 3.5 Flash-Lite, giảm chi phí cho các AI agent doanh nghiệp - Tin tức AI Aioga","description":"Google ra mắt Gemini 3.6 Flash và 3.5 Flash-Lite, nhằm giảm độ trễ và chi phí token của các tác nhân AI doanh nghiệp. Gemini 3.6 Flash tạo ra token ít hơn 17% so với thế hệ trước,...","url":"https://www.aioga.com/vi/news/cmruv44fg008zbii0eui6tcvw/","contentTranslated":true,"sourceHash":"57d4fef2ae1d2858","translatedAt":"2026-07-22T19:28:48.667Z"},"id":{"title":"Google merilis Gemini 3.6 Flash dan 3.5 Flash-Lite, mengurangi biaya agen AI perusahaan","summary":"Google merilis Gemini 3.6 Flash dan 3.5 Flash-Lite, bertujuan untuk mengurangi latensi dan biaya token AI agent perusahaan. Token keluaran 3.6 Flash berkurang 17% dibanding generasi sebelumnya, dengan tingkat keberhasilan 49% di DeepSWE (generasi sebelumnya 37%), dengan harga $1,50/1M token input.","category":"模型更新","source":"Artificial Intelligence News（RSS）","aggregationSource":"Artificial Intelligence News（RSS）","pageTitle":"Google merilis Gemini 3.6 Flash dan 3.5 Flash-Lite, mengurangi biaya agen AI perusahaan - Berita AI Aioga","description":"Google merilis Gemini 3.6 Flash dan 3.5 Flash-Lite, bertujuan untuk mengurangi latensi dan biaya token AI agent perusahaan. Token keluaran 3.6 Flash berkurang 17% dibanding generas...","url":"https://www.aioga.com/id/news/cmruv44fg008zbii0eui6tcvw/","contentTranslated":true,"sourceHash":"57d4fef2ae1d2858","translatedAt":"2026-07-22T19:29:36.525Z"},"th":{"title":"Google เปิดตัว Gemini 3.6 Flash และ 3.5 Flash-Lite เพื่อลดต้นทุนผู้ช่วย AI สำหรับธุรกิจ","summary":"Google เปิดตัว Gemini 3.6 Flash และ 3.5 Flash-Lite โดยมีเป้าหมายเพื่อลดความหน่วงและต้นทุน token ของ AI agent สำหรับองค์กร 3.6 Flash ลดจำนวน token ที่ส่งออกลง 17% เมื่อเทียบกับรุ่นก่อนหน้า มีอัตราความสำเร็จใน DeepSWE อยู่ที่ 49% (รุ่นก่อนหน้า 37%) ราคากำหนดอยู่ที่ $1.50 ต่อ 1M token ที่ป้อนเข้ามา","category":"模型更新","source":"Artificial Intelligence News（RSS）","aggregationSource":"Artificial Intelligence News（RSS）","pageTitle":"Google เปิดตัว Gemini 3.6 Flash และ 3.5 Flash-Lite เพื่อลดต้นทุนผู้ช่วย AI สำหรับธุรกิจ - ข่าว AI Aioga","description":"Google เปิดตัว Gemini 3.6 Flash และ 3.5 Flash-Lite โดยมีเป้าหมายเพื่อลดความหน่วงและต้นทุน token ของ AI agent สำหรับองค์กร 3.6 Flash ลดจำนวน token ที่ส่งออกลง 17% เมื่อเทียบกับรุ่นก...","url":"https://www.aioga.com/th/news/cmruv44fg008zbii0eui6tcvw/","contentTranslated":true,"sourceHash":"57d4fef2ae1d2858","translatedAt":"2026-07-22T19:29:43.269Z"},"pl":{"title":"Google wprowadza Gemini 3.6 Flash i 3.5 Flash-Lite, obniżając koszty inteligentnych agentów AI dla firm","summary":"Google wprowadziło Gemini 3.6 Flash i 3.5 Flash-Lite, mające na celu zmniejszenie opóźnień i kosztów tokenów dla inteligentnych agentów AI w przedsiębiorstwach. Gemini 3.6 Flash generuje o 17% mniej tokenów niż poprzednia generacja, osiągając 49% skuteczności w DeepSWE (poprzednia generacja 37%), a cena wynosi $1,50/1M tokenów wejściowych.","category":"模型更新","source":"Artificial Intelligence News（RSS）","aggregationSource":"Artificial Intelligence News（RSS）","pageTitle":"Google wprowadza Gemini 3.6 Flash i 3.5 Flash-Lite, obniżając koszty inteligentnych agentów AI dla firm - Aioga Wiadomości AI","description":"Google wprowadziło Gemini 3.6 Flash i 3.5 Flash-Lite, mające na celu zmniejszenie opóźnień i kosztów tokenów dla inteligentnych agentów AI w przedsiębiorstwach. Gemini 3.6 Flash ge...","url":"https://www.aioga.com/pl/news/cmruv44fg008zbii0eui6tcvw/","contentTranslated":true,"sourceHash":"57d4fef2ae1d2858","translatedAt":"2026-07-22T19:30:33.655Z"}}}}