Vertex AI Vector Search
cloud.google.com
- Free trial
- Web
- API
What Vertex AI Vector Search is
Vertex AI Vector Search is Google Cloud’s vector search engine for semantic search, recommendations, and generative AI applications. It uses Google Research’s ScaNN algorithm and supports dense embeddings, sparse embeddings, and hybrid search that combines semantic and keyword matching. Numeric and text-attribute filters can narrow results. Google describes uses including retrieval-augmented generation, chatbots, multimodal search, recommendations, and semantic search. An index can run on a public or private endpoint, including Private Service Connect, which provides private access to serving backends from multiple user VPC networks. Documented integrations include RAG Engine, Cloud Storage, Cloud Run, TensorFlow Recommenders, and LlamaIndex. Python and Java clients, a Python SDK, and REST workflows are documented. Security controls include VPC Service Controls and customer-managed encryption keys. The service is usage based: serving is charged per node hour, with separate index build, update, and streaming-update charges. Build and update data processing costs $3.00/GiB in all regions, while streaming inserts cost $0.45/GiB; serving rates vary by machine type and region. New Google Cloud customers may receive $300 in credits.
Who it suits
Vector Search suits developers building search, recommendation, retrieval-augmented generation, chatbot, or multimodal applications on Google Cloud. It also fits teams needing private endpoints and filtering.
Good to know, and what to watch
- Supports dense, sparse, and hybrid search
- Offers numeric and text metadata filtering
- Deploys to public or private endpoints
- Provides Python, Java, SDK, and REST options
- Includes Google Cloud security controls
- Usage-based pricing has separate charges
- Serving rates vary by machine type and region
- Regional quotas limit indexes and nodes
- Cloud deployment is required
Plans
Every plan and priceVertex AI Vector Search
The verdict
Vertex AI Vector Search provides flexible embedding search with hybrid matching, filtering, private connectivity, and Google Cloud integrations. Teams should account for separate usage charges and regional quotas when planning deployments.
What the maker publishes
- APIs and SDKs
- The documentation links Python and Java clients, the Python SDK, and REST API workflows. source
- Compliance
- Google’s compliance scope page lists Vector Search under Gemini Enterprise Agent Platform as under DISA review for DoD IL4 and IL5. source
- Endpoints
- Users can deploy an index to a public or private endpoint, including Private Service Connect configurations. source
- Filtering
- Searches can be restricted using numeric filtering and text attribute filtering. source
- Integrations
- Google documents use of Vector Search with RAG Engine, Cloud Storage, Cloud Run, TensorFlow Recommenders, and LlamaIndex. source
- Intended use
- Google describes use cases including semantic search, recommendations, retrieval-augmented generation, chatbots, and multimodal search. source
- Limits
- Per region and project, quotas include 5 concurrent index creations, 5 concurrent index updates, 50 deployed index nodes, and 100 indexes. source
- Private connectivity
- Private Service Connect supports private access to serving backends from multiple user VPC networks. source
- Purpose
- Vector Search is a vector search engine using Google Research’s ScaNN algorithm for search, recommendations, and generative AI applications. source
- Search modes
- It supports dense embeddings, sparse embeddings, and hybrid search combining semantic and keyword search. source
- Security
- Vector Search supports VPC Service Controls and customer-managed encryption keys, among other Google Cloud security controls. source
- Support
- Google provides support guidance for Vector Search and directs users to include diagnostic details when reporting problems. source
- Trial credits
- The overview page states that new Google Cloud customers can receive $300 in free credits to try Google Cloud products. source
Checked against the maker's pages on 8 Oct 2026.