{"@context":"https://schema.org","@type":"NewsArticle","generatedAt":"2026-09-22T18:00:52.144Z","headline":"Inception 发布扩散大语言模型 Mercury 2.5","description":"Inception 发布 Mercury 2.5，称其为目前能力最强、规模最大的扩散大语言模型，智能较 Mercury 2 提升 40%，可对标 GPT-5.6 Luna （Low）、Gemini 3.5 Flash-Lite 和 Claude Haiku 4.5。","url":"https://www.aioga.com/news/cmttautvn02p9rofpyocpqi9f/","mainEntityOfPage":"https://www.aioga.com/news/cmttautvn02p9rofpyocpqi9f/","datePublished":"2026-09-08T23:12:01.000Z","dateModified":"2026-09-08T23:12:01.000Z","inLanguage":"zh-CN","publisher":{"@type":"NewsMediaOrganization","name":"Aioga","url":"https://www.aioga.com"},"citation":["https://www.inceptionlabs.ai/blog/introducing-mercury-2-5","https://aihot.news/items/cmttautvn02p9rofpyocpqi9f"],"canonicalUrl":"https://www.aioga.com/news/cmttautvn02p9rofpyocpqi9f/","directAnswer":{"@type":"Answer","text":"Inception 发布 Mercury 2.5，并称其为该公司目前能力最强的生产模型及市场上能力最强的扩散大语言模型。发布材料称，其智能较 Mercury 2 提升40%。","url":"https://www.aioga.com/news/cmttautvn02p9rofpyocpqi9f/","dateCreated":"2026-09-08T23:12:01.000Z","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":"inceptionlabs.ai source article","url":"https://www.inceptionlabs.ai/blog/introducing-mercury-2-5","datePublished":"2026-09-08T23:12:01.000Z","provider":{"@type":"Organization","name":"inceptionlabs.ai","url":"https://www.inceptionlabs.ai/blog/introducing-mercury-2-5"}},{"@type":"CreativeWork","name":"AIHot archive record","url":"https://aihot.news/items/cmttautvn02p9rofpyocpqi9f","datePublished":"2026-09-08T23:12:01.000Z","provider":{"@type":"Organization","name":"AIHot","url":"https://aihot.news/items/cmttautvn02p9rofpyocpqi9f"}}],"aggregationSource":"Hacker News 热门（buzzing.cc 中文翻译）","originalPublisher":{"name":"inceptionlabs.ai","url":"https://www.inceptionlabs.ai/blog/introducing-mercury-2-5"},"geoDeepAnswer":null,"article":{"id":"cmttautvn02p9rofpyocpqi9f","slug":"cmttautvn02p9rofpyocpqi9f","url":"https://www.aioga.com/news/cmttautvn02p9rofpyocpqi9f/","title":"Inception 发布扩散大语言模型 Mercury 2.5","title_en":"","summary":"Inception 发布 Mercury 2.5，称其为目前能力最强、规模最大的扩散大语言模型，智能较 Mercury 2 提升 40%，可对标 GPT-5.6 Luna （Low）、Gemini 3.5 Flash-Lite 和 Claude Haiku 4.5。","source":"Hacker News 热门（buzzing.cc 中文翻译）","sourceUrl":"https://www.inceptionlabs.ai/blog/introducing-mercury-2-5","aiHotUrl":"https://aihot.news/items/cmttautvn02p9rofpyocpqi9f","publishedAt":"2026-09-08T23:12:01.000Z","category":"行业动态","score":58,"selected":false,"articleBody":["Today, we’re releasing Mercury 2.5, our most capable production model yet. It is a significant step up in quality over Mercury 2, with the same low-latency, low-cost serving profile.","Since Mercury 2’s launch, thousands of developers have built with it, dozens of enterprises have put it into production, and usage has grown over an order of magnitude. It now serves latency-sensitive workloads across search, voice, and coding products.","Those workloads gave us a clearer signal than benchmarks alone. We used customer feedback and production failure cases to sharpen the evals and focus training. Mercury 2.5 is the first result of that loop.","Mercury 2.5 is the most capable diffusion LLM on the market. To our knowledge, it is the largest diffusion language model ever trained.","Quality: 40% increase in intelligence from Mercury 2. Comparable to cost-optimized frontier models like GPT-5.6 Luna (Low), Gemini 3.5 Flash-Lite, and Claude Haiku 4.5.","Speed: 1,107 tokens per second on widely-available NVIDIA GPUs.","Price: $0.20 per million input and $0.75 per million output.","At launch, Mercury 2.5 is 80% off at $0.04 per million input and $0.15 per million output.","Capabilities: Tunable reasoning, parallel tool calls, and schema-aligned JSON.","Since Mercury 2's launch, we've watched Inception advance diffusion-based language models further on NVIDIA AI infrastructure. Mercury 2.5's step up in intelligence paired with sustained speeds and low costs, reflects how quickly new architectures can mature into production-ready systems on the NVIDIA platform.","Shruti Koparkar, Senior Manager of Product, Accelerated Computing Group at NVIDIA","One search request can trigger dozens of model calls: plan the search, rewrite queries, rerank results, structure facts, summarize sources, and check the answer. Mercury keeps those calls fast enough to stay inside a single user interaction. Several leading search-infrastructure companies now run it in production.","In voice, latency isn’t an infrastructure detail. It is the pause a caller hears.","OpenCall builds AI phone agents that handle live customer calls. On its production workload, Mercury brought median model response latency close to 170 milliseconds.","After we switched to Mercury, our P99 response time dropped from several minutes to just one second, and our P50 dropped from 0.4 seconds to under 0.2 — significantly faster than any other provider we’ve seen, and that’s including reasoning.","Oliver Silverstein, Co-founder and CEO, OpenCall","Read more: The first reasoning model fast enough to pick up the phone ：./mercury-2-the-first-reasoning-model-fast-enough-to-pick-up-the-phone","Coding agents already split work across models. One may plan or write code while others search, run tools, route requests, summarize state, or compact a long session. Those supporting calls happen again and again, so latency and cost compound quickly.","Augment Code uses Mercury for context compaction, model routing, and MCP tool search. Moving compaction to Mercury cut latency by 82%, from roughly 150 seconds to 27 seconds, and reduced cost by 90% while maintaining quality. Tool-search summaries return in under a second.","The same speed applies to developing web apps. Watch Mercury 2.5 generate a working music discovery log web app from a few prompts in the demo below.","Alongside Mercury 2.5, we’re announcing a preview of Mercury Voice：../models and Mercury Router：../models .","Mercury Voice delivers time-to-first-token (TTFT) under 170 milliseconds and is a dLLM optimized for voice agents with the tightest latency budgets.","Mercury Router understands incoming prompts with a dLLM and routes them to the best models (open and closed models) that offer the best mix of quality, speed, and cost.","Mercury models are available through our Inception API, Baseten, and OpenRouter. Enterprise deployments support dedicated capacity, autoscaling, compliance controls, and configurable data retention.","Try Mercury 2.5 in chat ：http://chat.inceptionlabs.ai/ Try the API with 100 million free tokens · Read the API docs ：https://docs.inceptionlabs.ai/get-started/get-started","Baseten customers: Deploy Mercury 2.5 through your existing Baseten setup.","Y Combinator companies: Claim：https://deals.ycombinator.com/deals/11140 $500,000 in deployment benefits.","Evaluating Mercury for voice? We’ll work with you to test workload fit, and validate performance under your serving constraints. Contact us：https://www.inceptionlabs.ai/enterprise#contact-sales .","We have already started training our next model. It is our largest model yet, and we are targeting a release in the coming months. Our next model will be a leap in capability without giving up diffusion’s speed and token-efficiency. That requires progress on model training, inference, evals, and infrastructure. If that's the kind of problem you want to work on, we’d love to hear from you：../careers .","Get Started：https://platform.inceptionlabs.ai/","Mercury Chat：https://chat.inceptionlabs.ai/","API Platform：https://platform.inceptionlabs.ai/","Documentation：https://docs.inceptionlabs.ai/get-started/get-started","Integrations：https://docs.inceptionlabs.ai/resources","Partners：../enterprise#customer-stories","Terms of Service：../docs/terms-of-use","Sales：mailto:sales@inceptionlabs.ai","Inquires：mailto:hello@inceptionlabs.ai","Discord：https://discord.com/invite/5VySp6ctXB","LinkedIn：https://www.linkedin.com/company/inception-labs-ai/"],"articleImages":[{"sourceUrl":"https://framerusercontent.com/images/RoZiv6hJStkiJmnDuDeCk9S2s.webp?width=512&height=341","alt":"","afterParagraph":0,"url":"/media/articles/cmtt8j0vv01iqrom8welx8v66/8cf446c324ffb21b.avif"},{"sourceUrl":"https://framerusercontent.com/images/BARaMxWJeRwFMQM349geFUJA5c8.png?lossless=1&width=1296&height=632","alt":"Speed Benchmark","afterParagraph":8,"url":"/media/articles/cmtt8j0vv01iqrom8welx8v66/b0b11dd60d46c9f2.webp"},{"sourceUrl":"https://framerusercontent.com/images/8lKo8OLi6Dq34kb2khNk8MwLo.png?width=2592&height=2780","alt":"Mercury 2.5 vs Mercury 2.0","afterParagraph":8,"url":"/media/articles/cmtt8j0vv01iqrom8welx8v66/271428869d8e7635.avif"},{"sourceUrl":"https://framerusercontent.com/images/Bs1DD6NdqSEgp8mzWLziwqm9yA.png?lossless=1&width=1296&height=924","alt":"Query Rewrite Latency Benchmark","afterParagraph":11,"url":"/media/articles/cmtt8j0vv01iqrom8welx8v66/1bc9a3418b1e6ac0.webp"},{"sourceUrl":"https://i.ytimg.com/vi_webp/NvR72NOcRPg/sddefault.webp","alt":"","afterParagraph":13,"url":"/media/articles/cmtt8j0vv01iqrom8welx8v66/0680c85adfe3204c.webp"},{"sourceUrl":"https://i.ytimg.com/vi_webp/PCPxLHM_NKE/sddefault.webp","alt":"","afterParagraph":19,"url":"/media/articles/cmtt8j0vv01iqrom8welx8v66/731c696c588b9491.webp"}],"mediaStatus":"ok","articleBodyZh":["今天，我们发布了 Mercury 2.5，这是我们迄今为止最强大的生产模型。与 Mercury 2 相比，它在质量上有显著提升，同时保持了低延迟、低成本的服务特点。","自 Mercury 2 发布以来，已有数千名开发者使用它进行开发，数十家企业将其投入生产，使用量增长了一个数量级以上。它现在为搜索、语音和编码产品提供对延迟敏感的工作负载服务。","这些工作负载给我们提供了比单独基准测试更清晰的信号。我们使用客户反馈和生产失败案例来优化评估并专注培训。Mercury 2.5 是这一循环的第一个成果。","Mercury 2.5 是市场上最强大的扩散语言大模型。据我们所知，它是迄今为止训练的最大扩散语言模型。","质量：相比 Mercury 2，智能度提升了 40%。可比拟于成本优化的前沿模型，如 GPT-5.6 Luna（低）、Gemini 3.5 Flash-Lite 以及 Claude Haiku 4.5。","速度：在广泛可用的 NVIDIA GPU 上达到每秒 1,107 个 token。","价格：每百万输入 $0.20，每百万输出 $0.75。","发布时，Mercury 2.5 提供 80% 折扣，每百万输入 $0.04，每百万输出 $0.15。","能力：可调推理、并行工具调用以及符合模式的 JSON。","自 Mercury 2 发布以来，我们见证了 Inception 在 NVIDIA AI 基础设施上推动基于扩散的语言模型进一步发展。Mercury 2.5 在保持高速度和低成本的同时提升的智能度，反映了新架构在 NVIDIA 平台上迅速成熟为生产就绪系统的能力。","Shruti Koparkar，NVIDIA 加速计算小组产品高级经理","一次搜索请求可能触发数十次模型调用：规划搜索、重写查询、重新排序结果、结构化事实、汇总来源并检查答案。Mercury 保持这些调用的速度足够快，以确保在单次用户交互内完成。多家领先的搜索基础设施公司现在已在生产中使用它。","在语音应用中，延迟不是基础设施细节，而是呼叫者听到的暂停。","OpenCall 构建了能处理实时客户通话的 AI 电话代理。在其生产工作负载中，Mercury 将模型响应的中位延迟降至接近 170 毫秒。","自从我们切换到 Mercury 后，我们的 P99 响应时间从几分钟降到了仅一秒钟，P50 从 0.4 秒降到不到 0.2 秒——比我们见过的任何其他提供商都快得多，这还包括推理时间。","奥利弗·西尔弗斯坦，联合创始人兼首席执行官，OpenCall","阅读更多：第一个足够快以接听电话的推理模型：./mercury-2-the-first-reasoning-model-fast-enough-to-pick-up-the-phone","编码代理已经将工作拆分到多个模型中。一个模型可能负责计划或编写代码，而其他模型则搜索、运行工具、路由请求、总结状态或压缩长会话。支持调用的操作会反复发生，因此延迟和成本会迅速累积。","Augment Code 使用 Mercury 进行上下文压缩、模型路由和 MCP 工具搜索。将压缩迁移到 Mercury 后，延迟降低了 82%，从大约 150 秒降至 27 秒，并在保持质量的同时将成本降低了 90%。工具搜索摘要在不到一秒钟内返回。","同样的速度也适用于开发 Web 应用。在下面的演示中，观看 Mercury 2.5 从几个提示生成一个可用的音乐发现日志 Web 应用。","随着 Mercury 2.5 的发布，我们还宣布 Mercury Voice：../models 和 Mercury Router：../models 的预览版。","Mercury Voice 提供低于 170 毫秒的首次生成标记时间 (TTFT)，是一种针对语音代理优化的 dLLM，具有最严格的延迟预算。","Mercury Router 使用 dLLM 理解输入提示，并将其路由到能够提供最佳质量、速度和成本组合的最佳模型（开放和封闭模型）。","Mercury 模型可通过我们的 Inception API、Baseten 和 OpenRouter 访问。企业部署支持专用容量、自动扩展、合规控制以及可配置的数据保留。","在聊天中试用 Mercury 2.5：http://chat.inceptionlabs.ai/ 使用 API 享受 1 亿免费代币 · 阅读 API 文档：https://docs.inceptionlabs.ai/get-started/get-started","Baseten 客户：通过您现有的 Baseten 设置部署 Mercury 2.5。","Y Combinator公司：声明：https://deals.ycombinator.com/deals/11140 提供50万美元的部署福利。","评估 Mercury 的语音功能？我们将与您合作测试工作负载的适配性，并验证在您的服务约束下的性能。联系我们：https://www.inceptionlabs.ai/enterprise#contact-sales。","我们已经开始训练我们的下一个模型。这是迄今为止最大规模的模型，我们计划在未来几个月发布。我们的下一个模型将在能力上实现飞跃，同时保持扩散模型的速度和令牌效率。这需要在模型训练、推理、评估和基础设施方面取得进展。如果这是您想要参与的类型问题，我们非常希望听到您的消息：../careers。","开始使用：https://platform.inceptionlabs.ai/","水星聊天：https://chat.inceptionlabs.ai/","API 平台：https://platform.inceptionlabs.ai/","文档：https://docs.inceptionlabs.ai/get-started/get-started","集成：https://docs.inceptionlabs.ai/resources","合作伙伴：../enterprise#customer-stories","服务条款：../docs/terms-of-use","销售：mailto:sales@inceptionlabs.ai","咨询：mailto:hello@inceptionlabs.ai","Discord：https://discord.com/invite/5VySp6ctXB","LinkedIn：https://www.linkedin.com/company/inception-labs-ai/"],"translationStatus":"translated","bodyOrigin":"source-page","editorial":{"summary":"Inception 发布 Mercury 2.5，并称其为该公司目前能力最强的生产模型及市场上能力最强的扩散大语言模型。发布材料称，其智能较 Mercury 2 提升40%。","background":"发布材料称，Mercury 2 已有数千名开发者使用、数十家企业投入生产，使用量增长超过一个数量级，并用于搜索、语音和编码产品的低延迟工作负载。","viewpoint":"Aioga 判断：此次发布主要强调质量提升，以及与 Mercury 2 相同的低延迟、低成本服务特征；“能力最强”“规模最大”和模型对标均应按发布方表述理解。","implications":"可能影响：对搜索、语音和编码产品中的低延迟工作负载而言，Mercury 2.5 可能成为待评估选项；发布方披露的质量、速度与价格不足以代表所有任务及运行条件。","nextStep":"后续观察：需要核对40%智能提升、每秒1,107个词元、输入输出价格及对标模型等披露，并关注这些结果在具体任务和硬件条件下的表现。","evidenceRefs":["title","summary","articleBody","source"],"status":"published","aiGenerated":true,"autoApproved":true,"generatedBy":"aioga-editorial:gpt-5.6-sol","reviewedBy":"aioga-editorial-review:gpt-5.6-sol","generatedAt":"2026-09-08T23:46:28.261Z","sourceHash":"67839737c42e2a6c","review":{"approved":true,"groundedness":97,"clarity":92,"duplicationRisk":12,"blockingIssues":[],"notes":["候选内容已将“能力最强”“规模最大”及模型对标明确归因于发布方，避免将厂商自述直接表述为独立事实。","“可能成为待评估选项”属于审慎的影响判断，并明确提示披露数据不代表所有任务及运行条件。"]},"validation":{"passed":true,"mode":"ai-auto","revisions":1,"checks":["schema","length","source-attribution","editorial-labels","inference-boundary","low-source-overlap","no-html","independent-ai-review"]}},"tags":["行业动态","Hacker News 热门（buzzing.cc 中文翻译）"],"translations":{"zh-CN":{"title":"Inception 发布扩散大语言模型 Mercury 2.5","summary":"Inception 发布 Mercury 2.5，称其为目前最强扩散大语言模型，智能较 Mercury 2 提升 40%，在 NVIDIA GPU 上达 1，107 tokens 每秒，上下文 260K tokens，定价 $0.20/百万输入、$0.75/百万输出，发布期 8 折优惠。","category":"行业动态","source":"inceptionlabs.ai","aggregationSource":"Hacker News：AI 热帖","pageTitle":"Inception 发布扩散大语言模型 Mercury 2.5 - Aioga AI资讯","description":"Inception 发布 Mercury 2.5，称其为目前最强扩散大语言模型，智能较 Mercury 2 提升 40%，在 NVIDIA GPU 上达 1，107 tokens 每秒，上下文 260K tokens，定价 $0.20/百万输入、$0.75/百万输出，发布期 8 折优惠。","url":"https://www.aioga.com/news/cmttautvn02p9rofpyocpqi9f/","articleBody":["今天，我们发布了 Mercury 2.5，这是我们迄今为止最强大的生产模型。与 Mercury 2 相比，它在质量上有显著提升，同时保持了低延迟、低成本的服务特点。","自 Mercury 2 发布以来，已有数千名开发者使用它进行开发，数十家企业将其投入生产，使用量增长了一个数量级以上。它现在为搜索、语音和编码产品提供对延迟敏感的工作负载服务。","这些工作负载给我们提供了比单独基准测试更清晰的信号。我们使用客户反馈和生产失败案例来优化评估并专注培训。Mercury 2.5 是这一循环的第一个成果。","Mercury 2.5 是市场上最强大的扩散语言大模型。据我们所知，它是迄今为止训练的最大扩散语言模型。","质量：相比 Mercury 2，智能度提升了 40%。可比拟于成本优化的前沿模型，如 GPT-5.6 Luna（低）、Gemini 3.5 Flash-Lite 以及 Claude Haiku 4.5。","速度：在广泛可用的 NVIDIA GPU 上达到每秒 1,107 个 token。","价格：每百万输入 $0.20，每百万输出 $0.75。","发布时，Mercury 2.5 提供 80% 折扣，每百万输入 $0.04，每百万输出 $0.15。","能力：可调推理、并行工具调用以及符合模式的 JSON。","自 Mercury 2 发布以来，我们见证了 Inception 在 NVIDIA AI 基础设施上推动基于扩散的语言模型进一步发展。Mercury 2.5 在保持高速度和低成本的同时提升的智能度，反映了新架构在 NVIDIA 平台上迅速成熟为生产就绪系统的能力。","Shruti Koparkar，NVIDIA 加速计算小组产品高级经理","一次搜索请求可能触发数十次模型调用：规划搜索、重写查询、重新排序结果、结构化事实、汇总来源并检查答案。Mercury 保持这些调用的速度足够快，以确保在单次用户交互内完成。多家领先的搜索基础设施公司现在已在生产中使用它。","在语音应用中，延迟不是基础设施细节，而是呼叫者听到的暂停。","OpenCall 构建了能处理实时客户通话的 AI 电话代理。在其生产工作负载中，Mercury 将模型响应的中位延迟降至接近 170 毫秒。","自从我们切换到 Mercury 后，我们的 P99 响应时间从几分钟降到了仅一秒钟，P50 从 0.4 秒降到不到 0.2 秒——比我们见过的任何其他提供商都快得多，这还包括推理时间。","奥利弗·西尔弗斯坦，联合创始人兼首席执行官，OpenCall","阅读更多：第一个足够快以接听电话的推理模型：./mercury-2-the-first-reasoning-model-fast-enough-to-pick-up-the-phone","编码代理已经将工作拆分到多个模型中。一个模型可能负责计划或编写代码，而其他模型则搜索、运行工具、路由请求、总结状态或压缩长会话。支持调用的操作会反复发生，因此延迟和成本会迅速累积。","Augment Code 使用 Mercury 进行上下文压缩、模型路由和 MCP 工具搜索。将压缩迁移到 Mercury 后，延迟降低了 82%，从大约 150 秒降至 27 秒，并在保持质量的同时将成本降低了 90%。工具搜索摘要在不到一秒钟内返回。","同样的速度也适用于开发 Web 应用。在下面的演示中，观看 Mercury 2.5 从几个提示生成一个可用的音乐发现日志 Web 应用。","随着 Mercury 2.5 的发布，我们还宣布 Mercury Voice：../models 和 Mercury Router：../models 的预览版。","Mercury Voice 提供低于 170 毫秒的首次生成标记时间 (TTFT)，是一种针对语音代理优化的 dLLM，具有最严格的延迟预算。","Mercury Router 使用 dLLM 理解输入提示，并将其路由到能够提供最佳质量、速度和成本组合的最佳模型（开放和封闭模型）。","Mercury 模型可通过我们的 Inception API、Baseten 和 OpenRouter 访问。企业部署支持专用容量、自动扩展、合规控制以及可配置的数据保留。","在聊天中试用 Mercury 2.5：http://chat.inceptionlabs.ai/ 使用 API 享受 1 亿免费代币 · 阅读 API 文档：https://docs.inceptionlabs.ai/get-started/get-started","Baseten 客户：通过您现有的 Baseten 设置部署 Mercury 2.5。","Y Combinator公司：声明：https://deals.ycombinator.com/deals/11140 提供50万美元的部署福利。","评估 Mercury 的语音功能？我们将与您合作测试工作负载的适配性，并验证在您的服务约束下的性能。联系我们：https://www.inceptionlabs.ai/enterprise#contact-sales。","我们已经开始训练我们的下一个模型。这是迄今为止最大规模的模型，我们计划在未来几个月发布。我们的下一个模型将在能力上实现飞跃，同时保持扩散模型的速度和令牌效率。这需要在模型训练、推理、评估和基础设施方面取得进展。如果这是您想要参与的类型问题，我们非常希望听到您的消息：../careers。","开始使用：https://platform.inceptionlabs.ai/","水星聊天：https://chat.inceptionlabs.ai/","API 平台：https://platform.inceptionlabs.ai/","文档：https://docs.inceptionlabs.ai/get-started/get-started","集成：https://docs.inceptionlabs.ai/resources","合作伙伴：../enterprise#customer-stories","服务条款：../docs/terms-of-use","销售：mailto:sales@inceptionlabs.ai","咨询：mailto:hello@inceptionlabs.ai","Discord：https://discord.com/invite/5VySp6ctXB","LinkedIn：https://www.linkedin.com/company/inception-labs-ai/"]},"en":{"title":"Inception Releases Diffusion Large Language Model Mercury 2.5","summary":"Inception released Mercury 2.5, claiming it to be the most capable and largest diffusion large language model to date, with a 40% intelligence improvement over Mercury 2. It can be compared to GPT-5.6 Luna (Low), Gemini 3.5 Flash-Lite, and Claude Haiku 4.5.","category":"Industry","source":"Hacker News 热门（buzzing.cc 中文翻译）","aggregationSource":"Hacker News：AI 热帖","pageTitle":"Inception Releases Diffusion Large Language Model Mercury 2.5 - Aioga AI News","description":"Inception released Mercury 2.5, claiming it to be the most capable and largest diffusion large language model to date, with a 40% intelligence improvement over Mercury 2. It can be...","url":"https://www.aioga.com/en/news/cmttautvn02p9rofpyocpqi9f/","contentTranslated":true,"sourceHash":"b4b4dea21f2687d4","translatedAt":"2026-09-08T23:41:13.776Z"},"ja":{"title":"Inception が拡散大規模言語モデル Mercury 2.5 を公開","summary":"Inception は Mercury 2.5 を公開し、これを現在最も能力が高く、規模の大きい拡散大規模言語モデルと称しています。Mercury 2 と比べて知性が40％向上しており、GPT-5.6 Luna（Low）、Gemini 3.5 Flash-Lite、Claude Haiku 4.5 と肩を並べることができます。","category":"業界動向","source":"Hacker News 热门（buzzing.cc 中文翻译）","aggregationSource":"Hacker News：AI 热帖","pageTitle":"Inception が拡散大規模言語モデル Mercury 2.5 を公開 - Aioga AIニュース","description":"Inception は Mercury 2.5 を公開し、これを現在最も能力が高く、規模の大きい拡散大規模言語モデルと称しています。Mercury 2 と比べて知性が40％向上しており、GPT-5.6 Luna（Low）、Gemini 3.5 Flash-Lite、Claude Haiku 4.5 と肩を並べることができます。","url":"https://www.aioga.com/ja/news/cmttautvn02p9rofpyocpqi9f/","contentTranslated":true,"sourceHash":"b4b4dea21f2687d4","translatedAt":"2026-09-08T23:41:14.939Z"},"ko":{"title":"Inception, 확산 대형 언어 모델 Mercury 2.5 발표","summary":"Inception은 Mercury 2.5를 발표하며, 이를 현재 가장 능력이 뛰어나고 규모가 큰 확산 대형 언어 모델이라고 소개했습니다. 지능 수준은 Mercury 2보다 40% 향상되었으며, GPT-5.6 Luna (Low), Gemini 3.5 Flash-Lite, Claude Haiku 4.5와 견줄 수 있습니다.","category":"업계 동향","source":"Hacker News 热门（buzzing.cc 中文翻译）","aggregationSource":"Hacker News：AI 热帖","pageTitle":"Inception, 확산 대형 언어 모델 Mercury 2.5 발표 - Aioga AI 뉴스","description":"Inception은 Mercury 2.5를 발표하며, 이를 현재 가장 능력이 뛰어나고 규모가 큰 확산 대형 언어 모델이라고 소개했습니다. 지능 수준은 Mercury 2보다 40% 향상되었으며, GPT-5.6 Luna (Low), Gemini 3.5 Flash-Lite, Claude Haiku 4.5와 견줄 수 있습니다.","url":"https://www.aioga.com/ko/news/cmttautvn02p9rofpyocpqi9f/","contentTranslated":true,"sourceHash":"b4b4dea21f2687d4","translatedAt":"2026-09-08T23:41:18.092Z"},"es":{"title":"Inception lanza el modelo de lenguaje difuso Mercury 2.5","summary":"Inception lanza Mercury 2.5, llamándolo hasta ahora el modelo de lenguaje difuso más potente y de mayor escala, con una inteligencia 40% superior a la de Mercury 2, comparable a GPT-5.6 Luna (Low), Gemini 3.5 Flash-Lite y Claude Haiku 4.5.","category":"Industria","source":"Hacker News 热门（buzzing.cc 中文翻译）","aggregationSource":"Hacker News：AI 热帖","pageTitle":"Inception lanza el modelo de lenguaje difuso Mercury 2.5 - Aioga Noticias de IA","description":"Inception lanza Mercury 2.5, llamándolo hasta ahora el modelo de lenguaje difuso más potente y de mayor escala, con una inteligencia 40% superior a la de Mercury 2, comparable a GP...","url":"https://www.aioga.com/es/news/cmttautvn02p9rofpyocpqi9f/","contentTranslated":true,"sourceHash":"b4b4dea21f2687d4","translatedAt":"2026-09-08T23:41:17.254Z"},"fr":{"title":"Inception publie le grand modèle de langage diffusif Mercury 2.5","summary":"Inception publie Mercury 2.5, le qualifiant de modèle de langage diffusif le plus puissant et le plus grand à ce jour, avec une intelligence améliorée de 40 % par rapport à Mercury 2, pouvant rivaliser avec GPT-5.6 Luna (Low), Gemini 3.5 Flash-Lite et Claude Haiku 4.5.","category":"Industrie","source":"Hacker News 热门（buzzing.cc 中文翻译）","aggregationSource":"Hacker News：AI 热帖","pageTitle":"Inception publie le grand modèle de langage diffusif Mercury 2.5 - Aioga Actualités IA","description":"Inception publie Mercury 2.5, le qualifiant de modèle de langage diffusif le plus puissant et le plus grand à ce jour, avec une intelligence améliorée de 40 % par rapport à Mercury...","url":"https://www.aioga.com/fr/news/cmttautvn02p9rofpyocpqi9f/","contentTranslated":true,"sourceHash":"b4b4dea21f2687d4","translatedAt":"2026-09-08T23:41:20.946Z"},"de":{"title":"Inception veröffentlicht das Diffusions-Sprachmodell Mercury 2.5","summary":"Inception veröffentlicht Mercury 2.5 und bezeichnet es als derzeit leistungsfähigstes und größtes Diffusions-Sprachmodell, seine Intelligenz ist gegenüber Mercury 2 um 40% verbessert und es kann mit GPT-5.6 Luna (Low), Gemini 3.5 Flash-Lite und Claude Haiku 4.5 konkurrieren.","category":"行业动态","source":"Hacker News 热门（buzzing.cc 中文翻译）","aggregationSource":"Hacker News：AI 热帖","pageTitle":"Inception veröffentlicht das Diffusions-Sprachmodell Mercury 2.5 - Aioga KI-News","description":"Inception veröffentlicht Mercury 2.5 und bezeichnet es als derzeit leistungsfähigstes und größtes Diffusions-Sprachmodell, seine Intelligenz ist gegenüber Mercury 2 um 40% verbesse...","url":"https://www.aioga.com/de/news/cmttautvn02p9rofpyocpqi9f/","contentTranslated":true,"sourceHash":"b4b4dea21f2687d4","translatedAt":"2026-09-08T23:41:20.706Z"},"pt-BR":{"title":"Inception lança grande modelo de linguagem de difusão Mercury 2.5","summary":"Inception lançou o Mercury 2.5, afirmando ser atualmente o modelo de linguagem de difusão mais potente e de maior escala, com inteligência 40% superior ao Mercury 2, podendo competir com GPT-5.6 Luna (Low), Gemini 3.5 Flash-Lite e Claude Haiku 4.5.","category":"行业动态","source":"Hacker News 热门（buzzing.cc 中文翻译）","aggregationSource":"Hacker News：AI 热帖","pageTitle":"Inception lança grande modelo de linguagem de difusão Mercury 2.5 - Aioga Notícias de IA","description":"Inception lançou o Mercury 2.5, afirmando ser atualmente o modelo de linguagem de difusão mais potente e de maior escala, com inteligência 40% superior ao Mercury 2, podendo compet...","url":"https://www.aioga.com/pt-BR/news/cmttautvn02p9rofpyocpqi9f/","contentTranslated":true,"sourceHash":"b4b4dea21f2687d4","translatedAt":"2026-09-08T23:41:24.136Z"},"ru":{"title":"Inception выпустила генеративную модель больших языков Mercury 2.5","summary":"Inception выпустила Mercury 2.5, заявив, что это самая мощная и крупная на данный момент генеративная модель больших языков; интеллект улучшен на 40% по сравнению с Mercury 2, может конкурировать с GPT-5.6 Luna (Low), Gemini 3.5 Flash-Lite и Claude Haiku 4.5.","category":"行业动态","source":"Hacker News 热门（buzzing.cc 中文翻译）","aggregationSource":"Hacker News：AI 热帖","pageTitle":"Inception выпустила генеративную модель больших языков Mercury 2.5 - Aioga Новости ИИ","description":"Inception выпустила Mercury 2.5, заявив, что это самая мощная и крупная на данный момент генеративная модель больших языков; интеллект улучшен на 40% по сравнению с Mercury 2, може...","url":"https://www.aioga.com/ru/news/cmttautvn02p9rofpyocpqi9f/","contentTranslated":true,"sourceHash":"b4b4dea21f2687d4","translatedAt":"2026-09-08T23:41:23.738Z"},"ar":{"title":"Inception تطلق نموذج اللغة الضخمة الانتشاري Mercury 2.5","summary":"أطلقت Inception Mercury 2.5، ووصفتها بأنها أقوى وأكبر نموذج لغة ضخمة انتشاري حاليًا، مع تحسين الذكاء بنسبة 40% مقارنة بـ Mercury 2، ويمكن مقارنتها مع GPT-5.6 Luna (Low)، Gemini 3.5 Flash-Lite وClaude Haiku 4.5.","category":"行业动态","source":"Hacker News 热门（buzzing.cc 中文翻译）","aggregationSource":"Hacker News：AI 热帖","pageTitle":"Inception تطلق نموذج اللغة الضخمة الانتشاري Mercury 2.5 - Aioga أخبار الذكاء الاصطناعي","description":"أطلقت Inception Mercury 2.5، ووصفتها بأنها أقوى وأكبر نموذج لغة ضخمة انتشاري حاليًا، مع تحسين الذكاء بنسبة 40% مقارنة بـ Mercury 2، ويمكن مقارنتها مع GPT-5.6 Luna (Low)، Gemini 3.5...","url":"https://www.aioga.com/ar/news/cmttautvn02p9rofpyocpqi9f/","contentTranslated":true,"sourceHash":"b4b4dea21f2687d4","translatedAt":"2026-09-08T23:41:27.284Z"},"hi":{"title":"Inception ने फैलाव बड़े भाषा मॉडल Mercury 2.5 जारी किया","summary":"Inception ने Mercury 2.5 जारी किया, जिसे अब तक की सबसे शक्तिशाली और सबसे बड़े फैलाव वाले बड़े भाषा मॉडल के रूप में वर्णित किया गया है। इसकी बुद्धिमत्ता Mercury 2 की तुलना में 40% बेहतर है और इसे GPT-5.6 Luna (Low), Gemini 3.5 Flash-Lite और Claude Haiku 4.5 के बराबर माना जा सकता है।","category":"行业动态","source":"Hacker News 热门（buzzing.cc 中文翻译）","aggregationSource":"Hacker News：AI 热帖","pageTitle":"Inception ने फैलाव बड़े भाषा मॉडल Mercury 2.5 जारी किया - Aioga AI समाचार","description":"Inception ने Mercury 2.5 जारी किया, जिसे अब तक की सबसे शक्तिशाली और सबसे बड़े फैलाव वाले बड़े भाषा मॉडल के रूप में वर्णित किया गया है। इसकी बुद्धिमत्ता Mercury 2 की तुलना में 40% ब...","url":"https://www.aioga.com/hi/news/cmttautvn02p9rofpyocpqi9f/","contentTranslated":true,"sourceHash":"b4b4dea21f2687d4","translatedAt":"2026-09-08T23:41:26.793Z"},"it":{"title":"Inception rilascia il modello di linguaggio diffuso Mercury 2.5","summary":"Inception ha rilasciato Mercury 2.5, definendolo il modello di linguaggio diffuso più potente e su larga scala finora, con un'intelligenza superiore del 40% rispetto a Mercury 2, comparabile a GPT-5.6 Luna (Low), Gemini 3.5 Flash-Lite e Claude Haiku 4.5.","category":"行业动态","source":"Hacker News 热门（buzzing.cc 中文翻译）","aggregationSource":"Hacker News：AI 热帖","pageTitle":"Inception rilascia il modello di linguaggio diffuso Mercury 2.5 - Aioga Notizie IA","description":"Inception ha rilasciato Mercury 2.5, definendolo il modello di linguaggio diffuso più potente e su larga scala finora, con un'intelligenza superiore del 40% rispetto a Mercury 2, c...","url":"https://www.aioga.com/it/news/cmttautvn02p9rofpyocpqi9f/","contentTranslated":true,"sourceHash":"b4b4dea21f2687d4","translatedAt":"2026-09-08T23:41:30.607Z"},"nl":{"title":"Inception lanceert diffusie-groottaalmodel Mercury 2.5","summary":"Inception lanceert Mercury 2.5, genoemd als het krachtigste en grootste diffusie-groottaalmodel tot nu toe, met een intelligentie die 40% hoger is dan Mercury 2, vergelijkbaar met GPT-5.6 Luna (Low), Gemini 3.5 Flash-Lite en Claude Haiku 4.5.","category":"行业动态","source":"Hacker News 热门（buzzing.cc 中文翻译）","aggregationSource":"Hacker News：AI 热帖","pageTitle":"Inception lanceert diffusie-groottaalmodel Mercury 2.5 - Aioga AI-nieuws","description":"Inception lanceert Mercury 2.5, genoemd als het krachtigste en grootste diffusie-groottaalmodel tot nu toe, met een intelligentie die 40% hoger is dan Mercury 2, vergelijkbaar met...","url":"https://www.aioga.com/nl/news/cmttautvn02p9rofpyocpqi9f/","contentTranslated":true,"sourceHash":"b4b4dea21f2687d4","translatedAt":"2026-09-08T23:41:29.689Z"},"tr":{"title":"Inception, Mercury 2.5 isimli büyük dil modelini yayımladı","summary":"Inception, Mercury 2.5'i yayımladı ve bunu şu anki en yetenekli ve en büyük yayılım (diffusion) tabanlı büyük dil modeli olarak tanımladı; Mercury 2'ye göre zekâ %40 artırılmıştır ve GPT-5.6 Luna (Low), Gemini 3.5 Flash-Lite ve Claude Haiku 4.5 ile karşılaştırılabilir.","category":"行业动态","source":"Hacker News 热门（buzzing.cc 中文翻译）","aggregationSource":"Hacker News：AI 热帖","pageTitle":"Inception, Mercury 2.5 isimli büyük dil modelini yayımladı - Aioga AI Haberleri","description":"Inception, Mercury 2.5'i yayımladı ve bunu şu anki en yetenekli ve en büyük yayılım (diffusion) tabanlı büyük dil modeli olarak tanımladı; Mercury 2'ye göre zekâ %40 artırılmıştır...","url":"https://www.aioga.com/tr/news/cmttautvn02p9rofpyocpqi9f/","contentTranslated":true,"sourceHash":"b4b4dea21f2687d4","translatedAt":"2026-09-08T23:41:33.269Z"},"vi":{"title":"Inception phát hành mô hình ngôn ngữ lớn khuếch tán Mercury 2.5","summary":"Inception phát hành Mercury 2.5, gọi đây là mô hình ngôn ngữ lớn khuếch tán mạnh nhất và quy mô lớn nhất hiện nay, trí thông minh tăng 40% so với Mercury 2, có thể so sánh với GPT-5.6 Luna (Low), Gemini 3.5 Flash-Lite và Claude Haiku 4.5.","category":"行业动态","source":"Hacker News 热门（buzzing.cc 中文翻译）","aggregationSource":"Hacker News：AI 热帖","pageTitle":"Inception phát hành mô hình ngôn ngữ lớn khuếch tán Mercury 2.5 - Tin tức AI Aioga","description":"Inception phát hành Mercury 2.5, gọi đây là mô hình ngôn ngữ lớn khuếch tán mạnh nhất và quy mô lớn nhất hiện nay, trí thông minh tăng 40% so với Mercury 2, có thể so sánh với GPT-...","url":"https://www.aioga.com/vi/news/cmttautvn02p9rofpyocpqi9f/","contentTranslated":true,"sourceHash":"b4b4dea21f2687d4","translatedAt":"2026-09-08T23:41:33.483Z"},"id":{"title":"Inception merilis model bahasa besar difusi Mercury 2.5","summary":"Inception merilis Mercury 2.5, menyebutnya sebagai model bahasa besar difusi dengan kemampuan terkuat dan ukuran terbesar saat ini, dengan peningkatan kecerdasan 40% dibanding Mercury 2, sejajar dengan GPT-5.6 Luna (Low), Gemini 3.5 Flash-Lite, dan Claude Haiku 4.5.","category":"行业动态","source":"Hacker News 热门（buzzing.cc 中文翻译）","aggregationSource":"Hacker News：AI 热帖","pageTitle":"Inception merilis model bahasa besar difusi Mercury 2.5 - Berita AI Aioga","description":"Inception merilis Mercury 2.5, menyebutnya sebagai model bahasa besar difusi dengan kemampuan terkuat dan ukuran terbesar saat ini, dengan peningkatan kecerdasan 40% dibanding Merc...","url":"https://www.aioga.com/id/news/cmttautvn02p9rofpyocpqi9f/","contentTranslated":true,"sourceHash":"b4b4dea21f2687d4","translatedAt":"2026-09-08T23:41:36.020Z"},"th":{"title":"Inception เปิดตัวโมเดลภาษาขนาดใหญ่แบบกระจาย Mercury 2.5","summary":"Inception เปิดตัว Mercury 2.5 โดยกล่าวว่าเป็นโมเดลภาษาขนาดใหญ่แบบกระจายที่มีความสามารถสูงสุดและใหญ่ที่สุดในปัจจุบัน มีสติปัญญาเกินกว่า Mercury 2 ถึง 40% สามารถเทียบได้กับ GPT-5.6 Luna (Low), Gemini 3.5 Flash-Lite และ Claude Haiku 4.5","category":"行业动态","source":"Hacker News 热门（buzzing.cc 中文翻译）","aggregationSource":"Hacker News：AI 热帖","pageTitle":"Inception เปิดตัวโมเดลภาษาขนาดใหญ่แบบกระจาย Mercury 2.5 - ข่าว AI Aioga","description":"Inception เปิดตัว Mercury 2.5 โดยกล่าวว่าเป็นโมเดลภาษาขนาดใหญ่แบบกระจายที่มีความสามารถสูงสุดและใหญ่ที่สุดในปัจจุบัน มีสติปัญญาเกินกว่า Mercury 2 ถึง 40% สามารถเทียบได้กับ GPT-5.6 L...","url":"https://www.aioga.com/th/news/cmttautvn02p9rofpyocpqi9f/","contentTranslated":true,"sourceHash":"b4b4dea21f2687d4","translatedAt":"2026-09-08T23:41:36.925Z"},"pl":{"title":"Inception wypuszcza rozproszony duży model językowy Mercury 2.5","summary":"Inception wypuszcza Mercury 2.5, określając go jako obecnie najpotężniejszy i największy rozproszony duży model językowy, z inteligencją wyższą o 40% w porównaniu z Mercury 2, mogący konkurować z GPT-5.6 Luna (Low), Gemini 3.5 Flash-Lite i Claude Haiku 4.5.","category":"行业动态","source":"Hacker News 热门（buzzing.cc 中文翻译）","aggregationSource":"Hacker News：AI 热帖","pageTitle":"Inception wypuszcza rozproszony duży model językowy Mercury 2.5 - Aioga Wiadomości AI","description":"Inception wypuszcza Mercury 2.5, określając go jako obecnie najpotężniejszy i największy rozproszony duży model językowy, z inteligencją wyższą o 40% w porównaniu z Mercury 2, mogą...","url":"https://www.aioga.com/pl/news/cmttautvn02p9rofpyocpqi9f/","contentTranslated":true,"sourceHash":"b4b4dea21f2687d4","translatedAt":"2026-09-08T23:41:40.938Z"}},"evidenceTier":"verified-news","reviewStatus":"automated-ingest","indexable":true,"editorialCover":""}}