제미나이 에이전트에서 클로드도 고를 수 있습니다

본문의 각주 번호는 아래 항목과 1:1 로 대응합니다. 정량 주장은 가능한 경우 둘 이상의 출처로 대조했고, 인용문은 원문에서 직접 확인한 것만 실었습니다.

1. 구글 클라우드는 2026년 10월 8일 업무용 범용 에이전트 「제미나이 에이전트」를 발표했고, 제미나이 계열과 앤트로픽 클로드 모델을 지금부터 쓰며 오픈·비공개 모델은 앞으로 더하고, 실시간 지출 상한과 에이전트별 고유 신원을 제공한다고 밝혔다Today at Gemini at Work 2026, Google Cloud announced the Gemini agent, a universal agent for work.https://blog.google/innovation-and-ai/infrastructure-and-cloud/google-cloud/gemini-at-work/확인 2026-10-12It runs each job on the model that fits best — orchestrating across our Gemini family of models and Claude models from Anthropic today, and other leading private and open models in the futurehttps://cloud.google.com/blog/products/ai-machine-learning/welcome-to-gemini-at-work-2026확인 2026-10-12Real-time spend caps: You can set a hard limit on a project's AI spend in the Cloud Billing Console, and Gemini enforces it.https://cloud.google.com/blog/products/ai-machine-learning/welcome-to-gemini-at-work-2026확인 2026-10-12Every agent gets its own identity, cryptographically attested and governed like an employee, with least-privilege permissions.https://cloud.google.com/blog/products/ai-machine-learning/welcome-to-gemini-at-work-2026확인 2026-10-12By default, the AI will pick the best model to complete the task, but users can also take over to choose a model — including those from third parties, starting with Anthropic's Claude models.https://techcrunch.com/2026/10/08/google-brings-agentic-ai-to-gemini-starting-with-businesses확인 2026-10-12
   ↩
2. 은행 고시 미국 달러 매매기준율이 1달러당 1,335.30원이었다(잡포스트의 하나은행 인용 · KB국민은행 192회차 14시 14분)하나은행이 제공한 환율 정보에 따르면 이날 미국 달러화의 매매기준율은 1달러당 1,335.30원으로 집계됐다.https://www.job-post.co.kr/news/articleView.html?idxno=231537확인 2026-10-12192 14:14:48 1,335.30https://obank.kbstar.com/quics?page=C101422&monyCd=USD확인 2026-10-12
   ↩
3. 구글은 2026년 10월 6일 이미지 생성·편집 모델 나노 바나나 2.1 을 정식 출시했고 글자 표현과 인포그래픽 배치가 나아졌다고 밝혔다Enhanced text rendering and infographic layout accuracyhttps://ai.google.dev/gemini-api/docs/models/gemini-nano-banana-2.1확인 2026-10-12구글이 현지 시각 10월 6일 제미나이 API 릴리스 노트를 통해 이미지 생성·편집 모델 '제미나이 나노 바나나 2.1(gemini-nano-banana-2.1)'을 정식 출시했다.https://zdnet.co.kr/view?no=20261007213423확인 2026-10-12
   ↩
4. 구글 가격표 기준 나노 바나나 2.1 의 이미지 출력 단가는 100만 토큰당 30달러로 나노 바나나 2의 60달러의 절반이고, 이미지 한 장은 1K 0.0336달러·2K 0.0504달러·4K 0.113달러(나노 바나나 2는 4K 0.151달러)여서 4K 는 약 25% 낮다Output price Not available $7.50 (text and thinking) $30.00 (images)https://ai.google.dev/gemini-api/docs/pricing확인 2026-10-12Equivalent to $0.0336 per 1K image*, $0.0504 per 2K image*, and $0.113 per 4K image*.https://ai.google.dev/gemini-api/docs/pricing확인 2026-10-12$60.00 (images) Equivalent to $0.045 per 0.5K image* $0.067 per 1K image*, $0.101 per 2K image*, and $0.151 per 4K image*.https://ai.google.dev/gemini-api/docs/pricing확인 2026-10-12
   ↩
5. 앤트로픽은 클로드 하이쿠 5.5 의 토큰 단가가 하이쿠 4.5 보다 10만 토큰 이하 요청에서 90% 낮고(입력 0.10달러·출력 0.50달러), 초과 요청은 50% 낮으며, 평균 실행 비용은 약 75% 줄었고, 복잡한 에이전트 코딩에는 소네트 5.5·오퍼스 5.5 가 더 맞는다고 밝혔다Claude Haiku 5.5 is priced 90% lower than Claude Haiku 4.5 for requests up to 100,000 tokens, and 50% lower for requests over 100,000 tokens. On Haiku 4.5, 90% of requests fell into the former category. This calculation also accounts for changes between Haiku 4.5 and Haiku 5.5 in how many tokens are used to complete a given piece of workhttps://www.anthropic.com/claude-haiku-5-5확인 2026-10-12On average, it now costs around 75% less to run.https://www.anthropic.com/claude-haiku-5-5확인 2026-10-12Sonnet 5.5 and Opus 5.5 remain better choices for complex agentic coding taskshttps://www.anthropic.com/claude-haiku-5-5확인 2026-10-1210만 토큰 이하 요청 시 입력/출력 토큰 가격이 100만 토큰당 각각 0.10달러, 0.50달러로 책정되어 전작 대비 무려 90% 인하됐다.https://aitimes.com/news/articleView.html?idxno=216044확인 2026-10-12Pricing for Claude Haiku 5.5 (Max) is $0.10 per 1M input tokens (competitively priced, median: $0.24) and $0.50 per 1M output tokens (competitively priced, median: $0.88).https://artificialanalysis.ai/models/claude-haiku-5-5확인 2026-10-12
   ↩
6. 독립 평가 기관 Artificial Analysis 는 클로드 하이쿠 5.5 의 지수를 43점, 출력 속도를 초당 236토큰으로 매겼고 평가 중 4억 4천만 토큰을 생성해 중앙값 1억 토큰보다 장황하다고 밝혔다Claude Haiku 5.5 (Max) scores 43 on the Artificial Analysis Intelligence Indexhttps://artificialanalysis.ai/models/claude-haiku-5-5확인 2026-10-12At 236 tokens per second, Claude Haiku 5.5 (Max) is notably fasthttps://artificialanalysis.ai/models/claude-haiku-5-5확인 2026-10-12it generated 440M tokens, which is very verbose in comparison to the median of 100M.https://artificialanalysis.ai/models/claude-haiku-5-5확인 2026-10-12
   ↩
7. 앤트로픽은 2026년 10월 8일 클로드 대시보드(유료 요금제 베타)와 클로드 모션(팀·엔터프라이즈 베타)을 공개했고, 차트마다 쿼리를 보여 주며, 문서·슬라이드·디자인은 무료 포함 모든 요금제에서 정식 제공한다고 밝혔다Claude Dashboards connects to a data platform like BigQuery, Databricks or Snowflake, or a CRM tool like Salesforce, and is in beta on paid plans. Claude Motion is in beta on Team and Enterprise. Claude Docs, Slides, and Design are out of beta and available on every plan, including Free.https://claude.com/resources/articles/dashboards-and-motion확인 2026-10-12Each chart shows the query behind it, and each animation is built as code you can edit.https://claude.com/resources/articles/dashboards-and-motion확인 2026-10-12Claude Dashboards, now in beta, lets users build live dashboards to see the exact queries pulling the data and export those dashboards to popular business intelligence tools.https://thenewstack.io/claude-dashboards-data-motion/확인 2026-10-12
   ↩
8. 앤트로픽은 클로드 대시보드를 BI 도구의 대체재로 내세우지 않고 기존 BI 도구와 함께 쓰도록 안내하며, 디일렉은 검증된 BI 플랫폼이 계속 쓰일 가능성이 있다고 봤다Dashboards work alongside your BI and analytics tools.https://claude.com/resources/articles/dashboards-and-motion확인 2026-10-12It’s worth noting that Anthropic isn’t positioning Dashboards as a replacement for business intelligence toolshttps://thenewstack.io/claude-dashboards-data-motion/확인 2026-10-12AI가 자동 생성한 대시보드에서 수치 정합성을 검증하기 위한 별도의 작업이 필요하다면 경영진 보고용이나 규제준수 대응을 위해선 검증된 BI 플랫폼이 계속 쓰일 가능성이 있다.https://thelec.kr/news/articleView.html?idxno=63510확인 2026-10-12
   ↩
9. 앤트로픽은 2026년 10월 6일(현지) 구글 독스·시트·슬라이드용 클로드를 모든 유료 요금제에 공개 베타로 열었고, 기본은 수정마다 승인하는 방식이며, 접근은 구글 공유 권한을 따르고, 엔터프라이즈 통제가 애드온에도 적용되며, 팀·엔터프라이즈는 소유자가 커넥터를 먼저 켜야 한다고 밝혔다Claude for Google Workspace™ is now in public beta on all paid Claude plans. It adds Claude to Google Docs, Sheets, and Slides, so that you can work with Claude directly in your open files.https://claude.com/resources/articles/claude-now-works-in-google-docs-sheets-and-slides확인 2026-10-12With the add-on, Claude opens in a sidebar next to your file, so you don’t need to switch between apps or browser tabs when you work.https://claude.com/resources/articles/claude-now-works-in-google-docs-sheets-and-slides확인 2026-10-12In the default “Ask before edits” mode, each change appears as an approval card with a summary and Claude waits for your approval to make the edit. In “Accept all edits” mode, Claude works through the task and applies changes without stopping.https://claude.com/resources/articles/claude-now-works-in-google-docs-sheets-and-slides확인 2026-10-12Claude’s access matches your existing Google sharing permissions.https://claude.com/resources/articles/claude-now-works-in-google-docs-sheets-and-slides확인 2026-10-12On Enterprise plans, controls such as the Compliance API, customer-managed encryption keys (CMEK), and OpenTelemetry audit export apply to the add-on toohttps://claude.com/resources/articles/claude-now-works-in-google-docs-sheets-and-slides확인 2026-10-12On Team and Enterprise plans, an owner or primary owner needs to enable the connectors first.https://claude.com/resources/articles/claude-now-works-in-google-docs-sheets-and-slides확인 2026-10-12Anthropic launched a public beta that integrates Claude into Google Docs, Sheets and Slides for all paid Claude plans today.https://www.reworked.co/digital-workplace/anthropic-opens-claude-for-google-workspace-in-beta/확인 2026-10-12
   ↩
10. 오픈AI 는 2026년 10월 1일 챗GPT 에 의류·액세서리 가상 착용과 즐겨찾기를 웹·모바일로 출시했고 기준 사진을 저장하며 설정에서 바꾸거나 지울 수 있으며, 테크크런치는 구글이 지난해 가상 착용을 냈고 오픈AI 의 즉시 결제는 성과가 좋지 않았다고 적었다Virtually try on clothing and accessories: A “Try on” button will appear on product listings for clothes and accessories in ChatGPT. Select it, then take or upload a selfie to generate a virtual try-on with ChatGPT Images. Your reference photo is saved for future try-ons. You can change or delete it in Settings → Personalization → Reference photos.https://help.openai.com/en/articles/6825453-chatgpt-release-notes확인 2026-10-12Save any product to Favorites or create a folder to organize your finds in your ChatGPT Library.https://help.openai.com/en/articles/6825453-chatgpt-release-notes확인 2026-10-12On Thursday, the company announced the global launch of two new shopping features, including a way to virtually try on clothing and accessories and a new favoriting functionhttps://techcrunch.com/2026/10/01/chatgpt-can-now-virtually-try-on-clothes-for-you/확인 2026-10-12OpenAI already had to pivot from one of its earlier ideas in this space, an instant checkout feature that ended up not performing well.https://techcrunch.com/2026/10/01/chatgpt-can-now-virtually-try-on-clothes-for-you/확인 2026-10-12Google launched virtual try-on last year.https://techcrunch.com/2026/10/01/chatgpt-can-now-virtually-try-on-clothes-for-you/확인 2026-10-12
    ↩
11. 오픈AI 는 2026년 10월 5일 앞으로 몇 주 안에 EU 의 챗GPT·코덱스 적격 텍스트 출력에 보이지 않는 워터마크를 추가한다고 밝혔고, 배경은 EU AI Act 요구이며, 자체 평가에서 단어 10% 교체 시 탐지율이 약 92%에서 66%, 25% 교체 시 17%였다고 밝혔다Over the coming weeks, we will add an invisible watermark to eligible ChatGPT and Codex text output in the European Union.https://openai.com/index/eu-text-provenance/확인 2026-10-12The EU AI Act requires generative AI providers to make generated text identifiable in a machine-readable way.https://openai.com/index/eu-text-provenance/확인 2026-10-12replacing 10% of words with synonyms reduced detection from about 92% to 66%. Replacing 25% of words reduced it to 17%.https://openai.com/index/eu-text-provenance/확인 2026-10-12OpenAI will begin automatically watermarking text generated with ChatGPT in the European Union, the company has announced.https://arstechnica.com/ai/2026/10/openai-will-watermark-chatgpt-outputs-by-default-but-only-in-the-eu확인 2026-10-12
    ↩
12. 마이크로소프트는 2026년 10월 7일 엔비디아 RTX Spark 기반 서피스 랩톱 울트라 선주문을 2,599달러부터 시작해 10월 16일 출하하며, 최대 6,144코어 블랙웰 RTX GPU·최대 20코어 그레이스 CPU·통합 메모리 최대 128GB·AI 성능 최대 1페타플롭을 밝혔고, 아르스 테크니카는 개발자용 서피스 RTX Spark 데브 박스가 5,999달러라고 전하며 보안과 사용자 수용 의문이 남았다고 적었다Starting today, pre-order Surface Laptop Ultra starting at $2,599 (MSRP) from Microsoft.com or select retailers worldwide, with availability beginning October 16.https://blogs.windows.com/devices/2026/10/07/pre-order-our-most-powerful-surface-devices-ever/확인 2026-10-12Surface Laptop Ultra unites an NVIDIA Blackwell RTX GPU with up to 6,144 cores, an NVIDIA Grace CPU with up to 20 cores, and up to 128 GB of unified memory.https://blogs.windows.com/devices/2026/10/07/pre-order-our-most-powerful-surface-devices-ever/확인 2026-10-12Based on theoretical FP4 performance of 1 petaflop using the sparsity feature.https://blogs.windows.com/devices/2026/10/07/pre-order-our-most-powerful-surface-devices-ever/확인 2026-10-12RTX Spark combines an NVIDIA Blackwell RTX GPU with up to 6,144 cores, and an up to 20-core NVIDIA Grace CPU connected at 600 GB/s.https://blogs.nvidia.com/blog/local-ai-rtx-spark-microsoft-windows-event확인 2026-10-12The Surface Laptop Ultra is now available on preorder and will begin shipping on Oct. 16https://cnbc.com/2026/10/07/microsoft-starts-taking-preorders-for-2599-surface-laptop-ultra.html확인 2026-10-12the company isn’t among the top six PC vendors in terms of unit shipments, according to an estimate from Gartnerhttps://cnbc.com/2026/10/07/microsoft-starts-taking-preorders-for-2599-surface-laptop-ultra.html확인 2026-10-12The device has a starting price of $2,599 and comes in various SoC and memory configurations, with either a 5120-core or 6144-core Blackwell GPU and up to 128GB of LPDDR5x unified memory. It will begin shipping on October 16 and is available for preorder now.https://arstechnica.com/gadgets/2026/10/microsoft-event-debuts-new-ai-friendly-hardware-and-windows-changes확인 2026-10-12This is a $5,999 anodized aluminum black box “designed to meet the needs of frontier developers.”https://arstechnica.com/gadgets/2026/10/microsoft-event-debuts-new-ai-friendly-hardware-and-windows-changes확인 2026-10-12questions about security and user acceptance still remainhttps://arstechnica.com/gadgets/2026/10/microsoft-event-debuts-new-ai-friendly-hardware-and-windows-changes확인 2026-10-12
    ↩
13. 마이크로소프트는 서피스 랩톱 울트라에서 파라미터 1,200억 개가 넘는 AI 모델을 로컬로 실행할 수 있다고 밝혔고, 엔비디아는 RTX Spark 가 엔비디아 AI 스택 전체를 윈도우 노트북과 소형 데스크톱에 넣는다고 밝혔다Run AI models exceeding 120B parameters locally, with up to 1 petaflop of AI performance.https://blogs.windows.com/devices/2026/10/07/pre-order-our-most-powerful-surface-devices-ever/확인 2026-10-12RTX Spark puts the full NVIDIA AI stack into Windows laptops and compact desktops, with laptop preorders open today and available October 16.https://blogs.nvidia.com/blog/local-ai-rtx-spark-microsoft-windows-event확인 2026-10-12
    ↩