[[["容易理解","easyToUnderstand","thumb-up"],["確實解決了我的問題","solvedMyProblem","thumb-up"],["其他","otherUp","thumb-up"]],[["難以理解","hardToUnderstand","thumb-down"],["資訊或程式碼範例有誤","incorrectInformationOrSampleCode","thumb-down"],["缺少我需要的資訊/範例","missingTheInformationSamplesINeed","thumb-down"],["翻譯問題","translationIssue","thumb-down"],["其他","otherDown","thumb-down"]],["上次更新時間:2025-09-04 (世界標準時間)。"],[],[],null,["# Model versions and lifecycle\n\nThis document defines key terms related to the lifecycle stages and important\ndates for Gemini and embedding models that are available on\nGoogle Cloud Vertex AI. It also gives you the recommended upgrades for\nthe models and points you to available migration paths.\n\nKey Terms\n---------\n\n**Stable model** : A publicly released version of the model that is available and\nsupported for production use starting on the release date. A stable model\nversion is typically released with a *retirement date*, which indicates the last\nday that the model is available. After this date, the model is no longer\naccessible or supported by Google.\n\n- **Latest stable model** : The latest version within the model family recommended for new and active projects and should be the target for migrations from earlier versions. See [Latest stable models](#latest-stable).\n- **Legacy stable model** : A model version that's been superseded by the Latest Stable Model. Although legacy stable models are still supported, you should strongly consider migrating to the latest model to receive the latest features and improvements. Access to legacy stable models might be restricted for new projects. See [Legacy stable models](#legacy-stable).\n\n**Retired model** : The model version is past its retirement date and has been\npermanently deactivated. Retired models are no longer accessible or supported by\nGoogle. API requests referencing a retired model ID typically returns a 404\nerror. See [Retired models](#retired-models).\n\n**Recommended upgrade** : The latest stable model that we recommend switching to.\nLatest stable models tend to offer better performance and more capabilities as\ncompared to legacy stable models. See the recommended upgrades in the\n[Legacy stable models](#legacy-stable) and [Retired models](#retired-models)\nsections.\n\nLatest stable models\n--------------------\n\nThe following table lists the latest stable models:\n\nLegacy stable models\n--------------------\n\nThe following table lists legacy stable models:\n\n\\*: Restricted for new projects.\n\nMigrate to a latest stable model\n--------------------------------\n\nTo learn how to migrate to a latest stable model, see\n[Migrate your application to Gemini 2 with the Gemini API in Vertex AI](/vertex-ai/generative-ai/docs/migrate-to-v2).\nThis guide gives you a set of migration steps that aims to minimize some\npotential risks involved in model migration and helps you use new models in an\noptimal way.\n\nHowever, if you don't have time to follow the guide and just need to quickly\nresolve the errors caused by models reaching their retirement dates, do the following:\n\n1. Update your application to point to the recommended upgrades.\n2. Test all mission critical features to make sure everything works as expected.\n3. Deploy the updates like you normally would.\n\nGemini auto-updated aliases\n---------------------------\n\nThe auto-updated alias of a Gemini model always points to the latest\nstable model. When a new latest stable model is available, the auto-updated\nalias automatically points to the new version.\n\nThe following table shows the auto-updated aliases for Gemini models\nand the latest stable models that they point to.\n\nRetired models\n--------------\n\n#### The following table lists the retired models (click to expand)\n\nWhat's next\n-----------\n\n- To learn which regions models are available, see [Deployments and endpoints](/vertex-ai/generative-ai/docs/learn/locations).\n- For details about individual models, see [Google models](/vertex-ai/generative-ai/docs/models) and click the model that you want to learn more about."]]