Hanzo SQL

Hanzo SQL is managed, serverless PostgreSQL 16. Provision a database in one call, connect with any Postgres client, and use instant branching, read replicas, connection pooling, and pgvector without running infrastructure.

Provision a Database

Databases are scoped to your organization. Creating one returns a ready-to-use connection string.

curl -X POST https://api.hanzo.ai/v1/sql \
  -H "Authorization: Bearer hk-..." \
  -H "X-Org-Id: org_a1b2c3" \
  -H "Content-Type: application/json" \
  -d '{ "name": "app", "region": "us-east" }'

The response contains a pooled endpoint (recommended for applications) and a direct endpoint (for migrations, LISTEN/NOTIFY, and advisory locks).

Connect

Any PostgreSQL driver or psql works — Hanzo SQL speaks the standard wire protocol over TLS.

psql "postgresql://user:password@db.hanzo.ai:6432/app?sslmode=require"

import asyncpg

conn = await asyncpg.connect(
    "postgresql://user:password@db.hanzo.ai:5432/app?sslmode=require"
)
rows = await conn.fetch("SELECT id, title FROM documents LIMIT 10")

Branches and Replicas

Branches are copy-on-write clones that share storage with the parent until data diverges — near-instant regardless of database size. Use them for development, CI, and schema previews.

# Branch the current state
curl -X POST https://api.hanzo.ai/v1/sql/app/branches \
  -H "Authorization: Bearer hk-..." \
  -d '{ "name": "feature-auth" }'

# Add a read replica
curl -X POST https://api.hanzo.ai/v1/sql/app/replicas \
  -H "Authorization: Bearer hk-..." \
  -d '{ "name": "app-ro-1" }'

Point read-heavy traffic at the replica endpoint; typical replication lag is under 100 ms.

Features

• PostgreSQL 16 with auto-scaling compute that scales to zero when idle
• pgvector for storing and searching embeddings alongside relational data
• Instant copy-on-write branches for dev, staging, and CI
• Read replicas, PgBouncer connection pooling, and point-in-time recovery
• Extensions: pg_trgm, postgis, uuid-ossp, pg_stat_statements, and more

Related

• Hanzo DB engine reference — pooling modes, PITR, and pgvector indexing
• Vector — a dedicated vector database for embedding-first workloads
• DocDB — MongoDB-compatible document storage
• API Keys — authenticate with hk- keys
• API Reference — every endpoint at api.hanzo.ai