Documentation
Cinch gives you isolated databases on demand. Each database scales to zero when idle and wakes automatically on the next request. You pay for what you use.
Connect in 30 seconds
pip install cinchdbfrom cinchdb import Cinch
import redis
# Create a database
client = Cinch(api_key="ck_live_...", org="acme")
scope = client.scope("my-app", environment="production", project="api")
db = scope.create_database(type="redis", name="cache")
# Connect with any Redis client
r = redis.from_url(str(db.url()))
r.set("hello", "world")
print(r.get("hello")) # b'world'Databases
Three database types, each accessible over TCP and HTTP. Create them through the dashboard, the SDK, or the REST API.
Redis
Cache, queues, search, vectors. Use any Redis client.
SQL
Full SQLite. Tables, joins, indexes, transactions.
Graph
Cypher queries. Use any Neo4j driver.
How it works
Create a database through the SDK, dashboard, or API. Each database is isolated, no shared resources, no noisy neighbors.
Connect with the client library you already use. Redis clients, SQL clients, Neo4j drivers, or plain HTTP.
Auto-stop: after 5 minutes idle, the database pauses and frees memory. The next request wakes it in ~10ms. After 24 hours idle, it archives to cold storage. Wake from archive takes ~25ms. All of this is transparent to your code.
Scopes control who can access what. Give each agent, tenant, or environment its own scope with its own databases and permissions. Learn more.
Platform SDKs
Use the SDK to create databases, manage scopes, and issue tokens. You connect to databases with standard client libraries (redis-py, neo4j, etc.), not the SDK.
Support
Questions? Email support@cinchdb.dev