Get started with CyborgDB in minutes! This guide will help you choose the right deployment model and get your first confidential vector search running quickly.

​
Choose Your Deployment Path

CyborgDB offers two main deployment approaches - both self-hosted on your infrastructure:

CyborgDB Service

Self-deployed REST API serviceDeploy as a microservice with REST API and client SDKs. Best for:
  • Multi-language team environments
  • Independent scaling capabilities
  • Operational simplicity with centralized management
  • Microservice architecture patterns

CyborgDB Embedded

Direct library integrationEmbed directly in your applications using Python/C++ libraries. Best for:
  • Maximum performance and control
  • Custom integrations and workflows
  • Single-application deployments
  • Sub-millisecond latency requirements

​
Framework Integrations

Already using AI frameworks? Get started even faster with our native integrations:

LangChain Integration

Drop-in replacement for existing vector storesReplace your current vector store with CyborgDB’s encrypted alternative:
from cyborgdb_core.integrations.langchain import CyborgVectorStore

store = CyborgVectorStore.from_texts(
    texts=["hello world", "goodbye world"],
    embedding="all-MiniLM-L6-v2",
    index_key=CyborgVectorStore.generate_key(),
    # ... other config
)

​
Quick Decision Guide

Fastest path to working confidential vector search→ Use CyborgDB Service with Python SDKDeploy as a microservice and connect via REST API. Minimal setup required.

​
Next Steps After Quickstart

Learn the Concepts

Understand the architectureDive deeper into how CyborgDB enables confidential vector search

Choose Your Database

PostgreSQL, Redis, or MemoryUnderstand database options and performance characteristics

Production Planning

Service vs Embedded comparisonReview deployment models and choose the right approach for production
Ready to get started? Choose your deployment path above and follow the step-by-step guide!