Skip to main content

Try CyborgDB in 5 minutes

Run this end-to-end — install, create an encrypted index, upsert, and query — with no server and no config:
Python
Store index_key before you upsert any data. CyborgDB has no key-recovery path — losing the key means losing the data. For anything past evaluation, keep the key in AWS Secrets Manager, Vault, or your KMS. See Managing Keys.
Without an API key the client runs in free-tier mode, capped at 1,000,000 items per index. Pass a key from the CyborgDB Admin Dashboard as the first argument (cyborgdb.Client(api_key, storage_config=...)) to lift the cap. See Get an API Key.

Deploying for real? Use the Service.

Embedded is the fastest path to a first query; the Service is what you deploy. It’s a self-hosted REST API with client SDKs for Python, JavaScript, Go, and REST.

Service Quickstart (Docker)

Spin up the REST service in one docker run and reach a first query.

Service Quickstart (Python)

Same REST service via pip install cyborgdb-service.

Migrating from another vector database

Already on Pinecone, Qdrant, Weaviate, ChromaDB, or Milvus? Bring your workload over with CyborgDB Migrate — an interactive TUI wizard (or headless TOML config) that resumes on interrupt and preserves IDs and metadata.

CyborgDB Migrate


Framework integrations

LangChain Integration

Drop-in replacement for existing vector stores:

More Integrations Coming

LlamaIndex, Haystack, Semantic Kernel, and custom frameworks. Request an integration →

Next steps

Learn the Concepts

How CyborgDB enables confidential vector search.

Choose a Backing Store

Memory, disk (default), or S3 — pick per environment.

Deployment Models

Embedded vs. Service, and when to pick each for production.