Google Vertex AI

Google Vertex AI is a fully-managed, unified AI development platform from Google Cloud for building, training, deploying, and managing machine learning and generative AI models at scale.

Features

• Managed Vector Search: Native support for vector search, including hybrid and semantic search for text, image, and other embeddings.
• Generative AI Models: Access to Gemini (multimodal), Imagen (image generation), Chirp (speech), Veo (video), and a wide variety of third-party and open-source models (e.g., Claude, Llama, Gemma) via Model Garden.
• Vertex AI Studio: Tool for prototyping, testing, and tuning generative AI models, with prompt design, evaluation, and rapid deployment.
• Custom Model Training: Train, test, and tune ML models using custom code, preferred ML frameworks, and advanced hyperparameter tuning.
• MLOps Tools: Integrated tools for automating, standardizing, and managing the ML lifecycle, including pipelines, feature store, model registry, and monitoring for skew/drift.
• Agent Builder: No-code and code-based tools for building and deploying generative AI agents grounded in your data.
• Notebook Integration: Vertex AI Notebooks (Colab Enterprise, Workbench) natively integrated with BigQuery and other Google Cloud services.
• Model Deployment: Support for batch or online predictions, prebuilt containers, and custom prediction routines.
• API Access: SDKs and APIs available for Python, Java, JavaScript, Go, and Curl.
• Data Handling: Support for image, video, text, tabular, and multimodal data for training and prediction.
• Scalability: Built on Google Cloud infrastructure for high availability, scalability, and security.
• Modular and Open: Open integration with OSS ML frameworks and other Google Cloud products.

Pricing

• Pay-as-you-go: Pricing is based on tools, storage, compute, and cloud resources used.
• Free Credits: New customers receive $300 in free credits to try Vertex AI and other Google Cloud products.
• Generative AI Models:
  • Imagen (image generation): Starting at $0.0001 per use
  • Text, chat, and code generation: Starting at $0.0001 per 1,000 characters
• AutoML Models:
  • Image data training: Starting at $1.375 per node hour
  • Video data training: Starting at $0.462 per node hour
  • Tabular data training: Contact sales
  • Text data training: Starting at $0.05 per hour
• Custom Models:
  • Custom model training: Price varies by machine type, region, and accelerators (contact sales)
• Vertex AI Notebooks:
  • Compute and storage charged per Google Compute Engine and Cloud Storage rates
• Management Fees: Apply based on usage (refer to Google Cloud pricing for details)
• Vertex AI Pipelines: Starting at $0.03 per pipeline run
• Vertex AI Vector Search: Serving and building costs depend on data size, queries per second (QPS), and nodes (refer to pricing example).

For full details and a pricing calculator, see Vertex AI Pricing.

Categories

• Vector Database Engines

Tags

managed-service, vector-search, hybrid-search, semantic-search, cloud-native