Vertex AI
Google Cloud's enterprise-grade machine learning platform, supporting Gemini API calls and using Service Account credentials for authentication. It is designed for production-level deployment.
Core Facts
Google provides API access to the Gemini model through two platforms, Google AI Studio and Vertex AI, with Google AI Studio offering developers a certain amount of free usage quota
Google AI Studio is an AI model debugging and prototype development platform launched by Google.
Vertex AI is Google Cloud's enterprise-grade machine learning platform, offering VPC network isolation, IAM permission management, and compliance auditing features.
Timeline (last 90 days)
The research node calls the Gemini model on Vertex AI via HTTP Request and mounts the Google Search tool to obtain real-time search results, thereby avoiding fabricating issues based solely on training data.
Image generation goes through Vertex AI, and the demo used the nano banana Pro model; the request body structure is content→role: user→parts→text; it returns Base64-encoded image data.
A Service Account is an identity type in Google Cloud used for machine-to-machine authentication; it has no password and proves its identity through a pair of keys (a unique email address and a private key); when calling the Vertex AI API, a personal account usually cannot be used for interactive authorization, so a service account is the standard solution.
When calling Vertex AI in n8n, authentication follows the Service Account route: in n8n's predefined credentials → Google Cloud API, fill in the region, the service account email, and the scope URL.
After applying for the Vertex AI free tier in the Google Cloud console, you can get about 300 dollars in credits.
当 BQML 与 Vertex AI 集成时，可以用 ML.PREDICT 做批量推理，或者通过 EXPORT MODEL 把模型导出到 GCS 供 Vertex AI Endpoint 使用。
使用 google-cloud-aiplatform>=1.30.0 时，它在 Dataset 资源里处理 bq_source 的方式与旧版本不同，可能导致 InvalidArgument 错误，因此应在容器镜像的 requirements.txt 中固定 SDK 版本。
在训练循环中运行复杂的大分区表 join 查询会与数仓争夺并发槽位，实际案例中 4 小时训练窗口可能被拉长到 12 小时。
通过 aiplatform.TabularDataset.create(display_name=..., bq_source=...) 创建的 BigqueryTabularDataset 是一个指向 BigQuery 表的指针，而不是数据副本。
许多团队将 BigQuery 等 Lakehouse 中的特征导出到新表或 GCS 桶供 Vertex AI 消费，导致数据重复存储，增加云成本。
26 more timeline events
All Facts (20)
Google provides API access to the Gemini model through two platforms, Google AI Studio and Vertex AI, with Google AI Studio offering developers a certain amount of free usage quota
90%VerifiedGoogle AI Studio is an AI model debugging and prototype development platform launched by Google.
90%VerifiedVertex AI is Google Cloud's enterprise-grade machine learning platform, offering VPC network isolation, IAM permission management, and compliance auditing features.
90%VerifiedOpenAI通过ChatGPT的插件系统和GPT Store构建AI Agent工具生态
75%UnverifiedGoogle AI付费版本通过Vertex AI或AI Studio提供更高配额，费用按Token消耗计算
85%Unverified谷歌云以BigQuery（数据分析）、Vertex AI（AI/ML）和GKE（Kubernetes容器编排）作为企业级市场的核心技术卖点
85%UnverifiedGoogle Cloud leverages its proprietary Gemini series models and Vertex AI platform to hold technical advantages in retrieval-augmented generation (RAG) and multimodal understanding.
70%UnverifiedGoogle AI Studio and Vertex AI are API platforms for developers that allow programmatic invocation of models and the setting of inference parameters such as temperature and top-p
60%UnverifiedAzure Machine Learning, AWS SageMaker, and Google Cloud Vertex AI form a tripartite landscape of cloud ML platforms.
60%UnverifiedImage generation goes through Vertex AI, and the demo used the nano banana Pro model; the request body structure is content→role: user→parts→text; it returns Base64-encoded image data.
50%UnverifiedThe research node calls the Gemini model on Vertex AI via HTTP Request and mounts the Google Search tool to obtain real-time search results, thereby avoiding fabricating issues based solely on training data.
50%UnverifiedAfter applying for the Vertex AI free tier in the Google Cloud console, you can get about 300 dollars in credits.
50%UnverifiedWhen calling Vertex AI in n8n, authentication follows the Service Account route: in n8n's predefined credentials → Google Cloud API, fill in the region, the service account email, and the scope URL.
50%UnverifiedA Service Account is an identity type in Google Cloud used for machine-to-machine authentication; it has no password and proves its identity through a pair of keys (a unique email address and a private key); when calling the Vertex AI API, a personal account usually cannot be used for interactive authorization, so a service account is the standard solution.
50%Unverified在训练循环中运行复杂的大分区表 join 查询会与数仓争夺并发槽位，实际案例中 4 小时训练窗口可能被拉长到 12 小时。
50%Unverified使用 google-cloud-aiplatform>=1.30.0 时，它在 Dataset 资源里处理 bq_source 的方式与旧版本不同，可能导致 InvalidArgument 错误，因此应在容器镜像的 requirements.txt 中固定 SDK 版本。
50%Unverified许多团队将 BigQuery 等 Lakehouse 中的特征导出到新表或 GCS 桶供 Vertex AI 消费，导致数据重复存储，增加云成本。
50%Unverified通过 aiplatform.TabularDataset.create(display_name=..., bq_source=...) 创建的 BigqueryTabularDataset 是一个指向 BigQuery 表的指针，而不是数据副本。
50%Unverified当 BQML 与 Vertex AI 集成时，可以用 ML.PREDICT 做批量推理，或者通过 EXPORT MODEL 把模型导出到 GCS 供 Vertex AI Endpoint 使用。
50%UnverifiedThe officially permitted uses of subscription-included API credits include: claude -p or the Agent SDK run directly with an API key, the Claude API and Console Playground; interactive Claude Code (terminal, IDE, desktop, web) as well as Bedrock, Vertex AI and Foundry are not available.
50%