Gemini for Google Cloud で使用される Gemini の大規模言語モデル（LLM）は、Gemini の基盤モデルのトレーニングに使用されるデータセットに加えて、一般公開されているコードのデータセット、 Google Cloud固有の資料、その他の関連する技術情報でトレーニングされます。モデルは、Gemini for Google Cloudのレスポンスが Google Cloud のユーザーにとって可能な限り有用になるようにトレーニングされています。
Gemini for Google Cloud は、基盤となるモデルのトレーニングやファインチューニングにプロンプトや生成されたレスポンスを使用しません。Gemini for Google Cloud モデルに送信されるデータは、リクエストへのレスポンスの提供のために使用が限定され、お客様から指示がない限り保存されません。Gemini for Google Cloudに送信される特定の情報をフィルタまたはブロックするために利用できるコントロールはありません。
Gemini for Google Cloud は、容量が許せばできるだけユーザーの近くでリクエストを処理します。Gemini forGoogle Cloudのサービス提供について詳しくは、Gemini for Google Cloud のロケーションをご覧ください。
Gemini for Google Cloud が情報源を引用する方法とタイミング
Gemini for Google Cloud の LLM は、他のスタンドアロンの LLM エクスペリエンスと同様に、オリジナル コンテンツを生成するものであり、既存のコンテンツを詳細に複製するものではありません。Google のシステムは、このような問題が発生する可能性を抑えるように設計されており、今後もこれらのシステムの動作は改善されます。
Gemini for Google Cloud がウェブページから詳細に直接引用している場合は、そのページを引用します。URL を含む回答の場合、Gemini forGoogle Cloud はユーザーによる表示を可能とし、場合によってはクリックしてソースページに直接移動できます。
[[["わかりやすい","easyToUnderstand","thumb-up"],["問題の解決に役立った","solvedMyProblem","thumb-up"],["その他","otherUp","thumb-up"]],[["わかりにくい","hardToUnderstand","thumb-down"],["情報またはサンプルコードが不正確","incorrectInformationOrSampleCode","thumb-down"],["必要な情報 / サンプルがない","missingTheInformationSamplesINeed","thumb-down"],["翻訳に関する問題","translationIssue","thumb-down"],["その他","otherDown","thumb-down"]],["最終更新日 2025-09-04 UTC。"],[[["\u003cp\u003eGemini for Google Cloud models are trained on publicly available code, Google Cloud-specific content, and other relevant technical data to provide responses that are highly useful for Google Cloud users.\u003c/p\u003e\n"],["\u003cp\u003eGemini for Google Cloud does not use user prompts or generated responses for training or fine-tuning its underlying models, and any data sent to it is used solely to serve a response and not stored unless directed.\u003c/p\u003e\n"],["\u003cp\u003eWhen Gemini for Google Cloud quotes directly from a webpage or code source, it will provide a citation to that source, including links where applicable, to increase transparency and user understanding of the sources.\u003c/p\u003e\n"],["\u003cp\u003eGoogle Cloud offers generative AI indemnification for Gemini for Google Cloud, assuming responsibility for potential legal risks if a user faces a copyright challenge after using content generated by the service.\u003c/p\u003e\n"],["\u003cp\u003eGemini Code Assist, a specific product within Gemini for Google Cloud, uses contextual information from files within your IDE and can be configured to avoid suggesting code that matches cited sources, while also offering code customization based on your organization's private codebase.\u003c/p\u003e\n"]]],[],null,["# How Gemini for Google Cloud works\n\nThe Gemini large language models (LLMs) that are used by Gemini for Google Cloud are trained on datasets of publicly available code, Google Cloud-specific material, and other relevant technical information in addition to the datasets used to train the Gemini [foundation models](https://storage.googleapis.com/deepmind-media/gemini/gemini_1_report.pdf). Models are trained so that Gemini for Google Cloud responses are as useful to Google Cloud users as possible.\n\n\u003cbr /\u003e\n\nAdditionally, Gemini in the Google Cloud console includes basic user\ncontextual information (project ID, product area, page title, and organization\nID) when giving responses. Google Cloud CLI (gcloud CLI) commands and\nother code may incorporate user context directly into the generated code\nsnippet. Gemini doesn't persist contextual information.\n\nGemini for Google Cloud doesn't use your prompts or\ngenerated responses for training or fine-tuning our underlying models. Any data\nsent to the Gemini for Google Cloud models is used strictly\nfor serving a response to the request, and unless instructed by you, isn't\nstored. There are no controls available to filter or block certain information\nfrom being sent to Gemini for Google Cloud.\n\nGemini for Google Cloud serves requests as close to the\nuser as possible where\ncapacity is available. For more information on where Gemini for\nGoogle Cloud\nserves from, see\n[Gemini for Google Cloud locations](/gemini/docs/locations).\n\nHow and when Gemini for Google Cloud cites sources\n--------------------------------------------------\n\nGemini for Google Cloud LLMs, like some other standalone\nLLM experiences, are intended to generate original content and not replicate\nexisting content at length. We've designed our systems to limit the chances of\nthis occurring, and we continue to improve how these systems function.\n\nIf Gemini for Google Cloud directly quotes at length from a\nweb page, it cites that page. For answers with URLs, Gemini for\nGoogle Cloud lets users see and, in some cases, click to navigate directly\nto the source page.\n\nWhen generating code or offering code completion, Gemini for\nGoogle Cloud provides citation information when it directly quotes at\nlength from another source, such as existing open source code. In the case of\ncitations to code repositories, the citation might also reference an applicable\nopen source license.\n\nHow Gemini for Google Cloud helps protect you with generative AI indemnification\n--------------------------------------------------------------------------------\n\nGemini for Google Cloud is covered as a\n[Generative AI Indemnified Service](https://cloud.google.com/terms/generative-ai-indemnified-services).\n\nIf you are challenged on copyright grounds after using content generated by\nGemini for Google Cloud, we assume certain responsibility\nfor the potential legal risks involved.\n\nFor full details about the indemnity, see our\n[Service Specific Terms](/terms/service-terms), or read our\n[blog post](https://cloud.google.com/blog/products/ai-machine-learning/protecting-customers-with-generative-ai-indemnification)\non this issue.\n\nGemini for Google Cloud products\n--------------------------------\n\nThe following sections provide additional details for specific\nGemini for Google Cloud products.\n\n### Gemini Code Assist\n\nTo allow for better code generation in IDEs,\n[Gemini Code Assist](/gemini/docs/codeassist/overview)\ngathers contextual information from the file that you're actively using in your\nIDE as well as other open and relevant local files in your project.\n\nWhen working with Gemini Code Assist in your IDE,\nGemini lists your project files (the context sources) that were\nused as reference to generate responses to your prompts. Context sources are\nshown every time you use Gemini chat.\n\nYou can prevent Gemini Code Assist from suggesting code that\nmatches cited sources by adjusting settings in your IDE\n([VS Code](/gemini/docs/codeassist/write-code-gemini#disable_code_suggestions_that_match_cited_sources),\n[Cloud Shell Editor](/code/docs/shell/write-code-gemini#disable_code_suggestions_that_match_cited_sources),\nand\n[Cloud Workstations](/workstations/docs/write-code-gemini#disable_code_suggestions_that_match_cited_sources)).\n\n[Code customization](/gemini/docs/codeassist/code-customization-overview) lets\nyou get code suggestions based on your organization's private codebase directly\nfrom Gemini Code Assist. To learn more about code customization,\nand how we provide security when accessing and storing your private code, see\nthe [Gemini Code Assist overview](/gemini/docs/codeassist/code-customization-overview).\nTo configure and use code customization, see\n[Configure and use Gemini Code Assist code customization](/gemini/docs/codeassist/code-customization).\n\nFor more information about Gemini Code Assist security controls,\nsee [Security, privacy, and compliance for Gemini Code Assist on\nGoogle Cloud](/gemini/docs/codeassist/security-privacy-compliance)."]]
