Cloud Vector Databases

Cloud vector databases provide scalable, managed vector storage solutions with enterprise-grade features and global availability.

Pinecone

Pinecone is a fully managed vector database service optimized for machine learning applications.

Usage

Pinecone provides:

• Fully managed vector storage

• Real-time indexing and search

• Horizontal scaling

• High-performance similarity search

• Enterprise security features

Inputs

Name	Display Name	Info
api_key	API Key	Pinecone API key for authentication
environment	Environment	Pinecone environment (e.g., us-east1-gcp)
index_name	Index Name	Name of the Pinecone index
namespace	Namespace	Optional namespace within the index

Outputs

Name	Display Name	Info
vector_store	Vector Store	Configured Pinecone vector store

Supabase

Supabase provides vector storage capabilities with PostgreSQL and pgvector extension.

Usage

Supabase vector features:

• PostgreSQL-based vector storage

• SQL query capabilities

• Real-time subscriptions

• Row-level security

• Built-in authentication

Inputs

Name	Display Name	Info
supabase_url	Supabase URL	Supabase project URL
supabase_key	API Key	Supabase API key
table_name	Table Name	Database table for vector storage
query_name	Query Name	Custom query function name

Outputs

Name	Display Name	Info
vector_store	Vector Store	Configured Supabase vector store

Upstash

Upstash provides serverless vector database services with Redis compatibility.

Usage

Upstash capabilities:

• Serverless vector storage

• Redis-compatible interface

• Global edge deployment

• Automatic scaling

• Pay-per-use pricing

Inputs

Name	Display Name	Info
url	Upstash URL	Upstash vector database URL
token	Token	Upstash authentication token
index_name	Index Name	Vector index name
namespace	Namespace	Optional namespace for organization

Outputs

Name	Display Name	Info
vector_store	Vector Store	Configured Upstash vector store

Vectara

Vectara offers a complete platform for neural search and retrieval-augmented generation.

Usage

Vectara platform features:

• Neural search capabilities

• Built-in embedding models

• Retrieval-augmented generation

• Multi-modal search

• Enterprise security

Inputs

Name	Display Name	Info
customer_id	Customer ID	Vectara customer identifier
corpus_id	Corpus ID	Vectara corpus identifier
api_key	API Key	Vectara API key

Outputs

Name	Display Name	Info
vector_store	Vector Store	Configured Vectara vector store

Vectara Search

Specialized search component for Vectara with advanced query capabilities.

Usage

Advanced Vectara search:

• Semantic search queries

• Hybrid search capabilities

• Result re-ranking

• Query expansion

• Multi-corpus search

Inputs

Name	Display Name	Info
search_query	Search Query	Natural language search query
num_results	Results Count	Number of results to return
metadata_filter	Metadata Filter	Optional metadata filtering

Outputs

Name	Display Name	Info
search_results	Search Results	Ranked search results

Usage Notes

• Managed Infrastructure: Cloud providers handle scaling and maintenance

• Global Availability: Multi-region deployment options

• Enterprise Features: Advanced security, compliance, and monitoring

• Performance Optimization: Built-in performance tuning and optimization

• API Integration: RESTful APIs and SDK support

• Cost Efficiency: Pay-as-you-scale pricing models