Google Gemini / Vertex AI

Vertex AI Persistent Resources GA and Workbench M129 Release

Vertex AI persistent resources for custom training is now generally available with reboot support. Vertex AI Workbench instances receive M129 update including upgraded Dataproc JupyterLab plugin to version 0.1.85. These are feature enhancements with no breaking changes or required actions.

Date not specified

Google releases Llama 4 Maverick and Scout models in preview on Vertex AI

Google has released managed APIs for Llama 4 Maverick and Scout models in preview on Vertex AI. This expands the available model options for developers using Vertex AI's generative AI platform. No immediate action is required as this is a new capability addition.

Date not specified

Google Vertex AI releases Gemini 2.5 Pro and new features

Google has released several new capabilities across Vertex AI including Gemini 2.5 Pro in preview, enhanced Gemini Live API with multi-voice support and screen sharing, and the new Agent Development Kit. Additional features include Agent Garden, improved grounding capabilities with Google Maps, and rebranding of Agent Builder components. These are primarily new feature releases that expand platform capabilities without requiring immediate user action.

Date not specified

Colab Enterprise adds default GPU runtime switching feature in preview

Google Colab Enterprise now allows users to switch to a default runtime with GPUs using a button in notebooks. This is a preview feature that requires administrator configuration to enable GPU access for users. No immediate action required unless organizations want to enable this capability.

Date not specified

Vertex AI now supports Spot VMs for training and prediction jobs

Google has made Spot VMs generally available for Vertex AI training and prediction workloads, offering significant cost discounts on excess Compute Engine capacity. Users can now opt to use these preemptible instances to reduce operational costs, though Google may reclaim capacity at any time. This is an optional cost optimization feature that doesn't require any immediate action.

Date not specified

Vertex AI adds GPU reservations and Workbench backup/restore

Two new features are now generally available: GPU-attached VM reservations for training and prediction jobs, and backup/restore functionality for Vertex AI Workbench instances. These enhancements improve resource availability assurance and data protection capabilities for users.

Date not specified

New AI Models Available in Vertex AI Model Garden

Google has added DeepSeek-V3-0324, TxGemma, and Sesame CSM models to Model Garden. DeepSeek-R1, V3, and V3-0324 models can now be deployed with H200 GPUs and improved vLLM support. Deployment is available through Python SDK, gcloud CLI, or API in Preview mode.

Date not specified

Anthropic Claude Sonnet 3.7 GA on Vertex AI with Confidential Computing preview

Anthropic's Claude Sonnet 3.7 model is now generally available on Vertex AI with Provision Throughput support. Additionally, Vertex AI Workbench introduces Confidential Computing in preview, allowing users to encrypt data-in-use when creating instances.

Date not specified

Google Vertex AI adds multimodal capabilities to Mistral Small 3.1

Google Vertex AI has added Mistral Small 3.1 (25.03) to Model Garden with new multimodal capabilities and extended context window of up to 128,000 tokens. This expands the available model options for developers using Vertex AI. No action required - this is a new capability addition.

Date not specified

Google releases Judge model evaluation and customization tools in Preview for Gen AI evaluation service

Google has released Judge model evaluation and customization tools in Preview for the Gen AI evaluation service in Vertex AI. This new capability allows users to evaluate and customize generative AI models. No immediate action required as this is a new feature addition in preview status.

Date not specified

Gemini Context Caching Reaches General Availability on Vertex AI

Context caching functionality for Gemini models on Vertex AI is now generally available, moving from preview to production-ready status. This feature allows for improved performance and cost efficiency when processing repeated context. No action required, existing implementations should continue working normally.

Date not specified

Google Vertex AI adds new models (Gemma 3, ShieldGemma 2, CogVideoX-2b) and updates fine-tuning tools

Google has released new AI models in Model Garden including Gemma 3, ShieldGemma 2, and CogVideoX-2b, along with enhanced fine-tuning capabilities for Llama 3.1 and Gemma 2. The M128 release also brings miscellaneous package updates across Vertex AI Workbench components. These are additive capabilities that expand available model options and development tools.

Date not specified

Google Gemini 2.0 Flash fine-tuning reaches GA with function calling support

Google has made Gemini 2.0 Flash fine-tuning generally available on Vertex AI, moving from preview to production-ready status. The release includes new support for tuning function calling capabilities, expanding customization options for developers.

Date not specified

Vertex AI Agent Engine GA with Billing Starting March 4, 2025

Google's Vertex AI Agent Engine has reached general availability and billing begins March 4, 2025. The service was previously known as LangChain on Vertex AI and has been renamed. Users should delete unused resources to avoid unwanted charges as the free preview period ends.

Date not specified

Google Colab Enterprise adds Terraform integration for notebook management

Google Colab Enterprise now supports Terraform resources for scheduling notebook runs and managing runtimes and runtime templates. This adds infrastructure-as-code capabilities for enterprise users to programmatically manage their notebook environments. No action required, this is an additive feature enhancement.

Date not specified

Google releases Gemini 2.0 Flash-Lite on Vertex AI

Google has released Gemini 2.0 Flash-Lite as a generally available model on Vertex AI. This represents a new model option for users of the platform. No immediate action is required as this is an additive capability enhancement.

Date not specified

Anthropic Claude Sonnet 3.7 now available in Preview on Vertex AI

Google has added Anthropic's Claude Sonnet 3.7 model to Vertex AI in preview mode. Users can access this new model through Model Garden with associated model card documentation. This expands the available AI model options on the Vertex AI platform.

Date not specified

Vertex AI Adds New Evaluation Metrics, Model Updates, and Enhanced Fine-tuning

Google Vertex AI introduced new evaluation metrics (perplexity, BLEU, ROUGE variants), service account selection for fine-tuning, and updated support for Llama 3.1, Gemma 2, PaliGemma 2, and LLaVA Next models. The updates include new tutorial notebooks and improved fine-tuning workflows with PEFT and Axolotl frameworks. These are additive features that enhance existing capabilities without requiring immediate action.

Date not specified

Google Vertex AI adds DeepSeek-V3 and DeepSeek-R1 models to Model Garden

Google has added two new DeepSeek models to Vertex AI Model Garden: DeepSeek-V3 (671B parameter MoE language model) and DeepSeek-R1 (671B parameter reasoning model comparable to OpenAI-o1). Both models are available in Preview mode and can be deployed using provided notebooks. This expands the available model options for developers using Vertex AI's Model Garden platform.

Date not specified

Google releases Llama 3.3 70B model in Preview on Vertex AI

Google has released the Llama 3.3 70B model in Preview status on Vertex AI's managed platform. This provides users with access to a new large language model option for their AI applications. No immediate action is required as this is a new capability addition.

Date not specified