在 Google Play 服務 API 中使用 LiteRT 時,輸入資料 (例如圖片、影片、文字) 的處理作業會完全在裝置上執行,Google Play 服務 API 中的 TensorFlow Lite 不會將該資料傳送至 Google 伺服器。因此,您可以使用我們的 API 處理不應離開裝置的資料。
Google Play 服務 API 中的 LiteRT 可能會不時與 Google 伺服器連線,以便接收錯誤修正、更新模型和硬體加速器相容性資訊等內容。Google Play 服務 API 中的 LiteRT 也會將應用程式中 API 的效能和使用率指標傳送給 Google。Google 會使用這些指標資料來評估效能、偵錯、維護及改善 API,並偵測濫用或濫用行為,詳情請參閱我們的《隱私權政策》。
您有責任依據適用法律規定,向應用程式使用者說明 Google 如何處理 Google Play 服務 API 指標資料中的 LiteRT。
我們收集的資料包括:
裝置資訊 (例如製造商、型號、OS 版本和版本) 和可用的 ML 硬體加速器 (GPU 和 DSP)。用於診斷和使用情形分析。
[[["容易理解","easyToUnderstand","thumb-up"],["確實解決了我的問題","solvedMyProblem","thumb-up"],["其他","otherUp","thumb-up"]],[["缺少我需要的資訊","missingTheInformationINeed","thumb-down"],["過於複雜/步驟過多","tooComplicatedTooManySteps","thumb-down"],["過時","outOfDate","thumb-down"],["翻譯問題","translationIssue","thumb-down"],["示例/程式碼問題","samplesCodeIssue","thumb-down"],["其他","otherDown","thumb-down"]],["上次更新時間:2025-07-24 (世界標準時間)。"],[],[],null,["# LiteRT in Google Play services\n\nLiteRT is available in Google Play services runtime for all Android\ndevices running the current version of Play services. This runtime allows you to\nrun machine learning (ML) models without statically bundling LiteRT\nlibraries into your app.\n\nWith the Google Play services API, you can reduce the size of your apps and gain\nimproved performance from the latest stable version of the libraries. TensorFlow\nLite in Google Play services is the recommended way to use LiteRT on\nAndroid.\n\nIf you are already using standalone TensorFlow Lite in your app, refer to the\n[Migrating from standalone LiteRT](./java#migrating) section to update an\nexisting app to use the Play services runtime. For more information about Google\nPlay services, see the [Google Play\nservices](https://developers.google.com/android/guides/overview) website.\n| **Terms:** By accessing or using LiteRT in Google Play services APIs, you agree to the [Terms of Service](#tos). Please read and understand all applicable terms and policies before accessing the APIs.\n\nUsing the Play services runtime\n-------------------------------\n\nThe LiteRT in Google Play services is available through the following\nprogramming language apis:\n\n- Java API - [see guide](./java)\n- C/C++ APIs - [see guide](./native)\n\n| **Note:** LiteRT in Google Play services uses the `play-services-tflite-java` package, for both Java and C/C++.\n\nLimitations\n-----------\n\nLiteRT in Google Play services has the following limitations:\n\n- Support for hardware acceleration delegates is limited to the delegates listed in the [Hardware acceleration](../java#hardware-acceleration) section. No other acceleration delegates are supported.\n- Experimental or deprecated LiteRT APIs, including custom ops, are not supported.\n\nSupport and feedback\n--------------------\n\nYou can provide feedback and get support through the TensorFlow Issue Tracker.\nPlease report issues and support requests using the\n[Issue template](https://github.com/tensorflow/tensorflow/issues/new?title=TensorFlow+Lite+in+Play+Services+issue&template=tflite-in-play-services.md)\nfor LiteRT in Google Play services.\n\nTerms of service\n----------------\n\nUse of LiteRT in Google Play services APIs is subject to the\n[Google APIs Terms of Service](https://developers.google.com/terms/).\n\n### Privacy and data collection\n\nWhen you use LiteRT in Google Play services APIs, processing of the\ninput data, such as images, video, text, fully happens on-device, and TensorFlow\nLite in Google Play services APIs does not send that data to Google servers. As\na result, you can use our APIs for processing data that should not leave the\ndevice.\n\nThe LiteRT in Google Play services APIs may contact Google servers from\ntime to time in order to receive things like bug fixes, updated models and\nhardware accelerator compatibility information. The LiteRT in Google\nPlay services APIs also sends metrics about the performance and utilization of\nthe APIs in your app to Google. Google uses this metrics data to measure\nperformance, debug, maintain and improve the APIs, and detect misuse or abuse,\nas further described in our\n[Privacy Policy](https://policies.google.com/privacy).\n\n**You are responsible for informing users of your app about Google's processing\nof LiteRT in Google Play services APIs metrics data as required by\napplicable law.**\n\nData we collect includes the following:\n\n- Device information (such as manufacturer, model, OS version and build) and available ML hardware accelerators (GPU and DSP). Used for diagnostics and usage analytics.\n- Device identifier used for diagnostics and usage analytics.\n- App information (package name, app version). Used for diagnostics and usage analytics.\n- API configuration (such as which delegates are being used). Used for diagnostics and usage analytics.\n- Event type (such as interpreter creation, inference). Used for diagnostics and usage analytics.\n- Error codes. Used for diagnostics.\n- Performance metrics. Used for diagnostics."]]