Serverless ClickHouse

Columnar analytics at production scale, on a connection string. Web SQL console, native and HTTP endpoints, the same tools you already use.

Create ClickHouse Create an account

SQL Console

ClickHouse

Tables

users
events
orders
sessions

SELECT id, name, created_at

FROM events

WHERE created_at > NOW() - INTERVAL '7 days'

ORDER BY id DESC LIMIT 5;

id	name	created_at
1247	signup	2026-03-12 14:21
1246	purchase	2026-03-12 14:08
1245	login	2026-03-12 13:55
1244	signup	2026-03-12 13:44
1243	purchase	2026-03-12 13:30

5 rows · 12ms

Where to run it

Cloud, desktop, or self-host

One database, three places it can live. Pick the speed you want and move between them when you outgrow it.

Layerbase Cloud

ClickHouse provisioned on managed infrastructure.

Create database

Layerbase Desktop

Run ClickHouse locally with the same dashboard and query console.

Download

SpinDB

Open-source CLI to run ClickHouse on your own infra.

View on GitHub

What you get

Production features, free tier included

Every ClickHouse database ships with the same primitives: TLS, web console, hibernation, and a real dashboard. No tiered features hidden behind a sales call.

Web IDE

Query console in the dashboard, no extra client to install.

TLS by default

All connections encrypted. No setup required.

Always-on connections

Optional keep-alive add-on prevents hibernation.

Scale to zero

Hibernates when idle. Wakes on the next connection.

Connection pooling

Pooled endpoint for serverless and edge workloads.

Direct connections

Bypass the pooler for migrations and replication.

HTTPS access

HTTP-native engines reachable over port 443.

Serverless driver

Upstash REST or PlanetScale serverless drivers supported.

IP whitelisting

Restrict access to specific IPs or CIDR ranges.

Automatic backups

Scheduled point-in-time backups with one-click restore.

Quick start

Connect with any ClickHouse client

Real-time analytics, time-series, observability backends, ad-hoc OLAP queries over large datasets, and apps that outgrew Postgres for reporting.

Stand up a ClickHouse for a single project without provisioning a cluster
Replicate slow Postgres reporting queries to a columnar engine
Build an observability backend (logs, metrics) on the same platform as your operational databases
Prototype real-time analytics dashboards on the free tier
Self-host on your own hardware with SpinDB for tighter compliance control

bashClickHouse

curl -u layerbase:password "https://your-host.cloud.layerbase.dev/?query=SELECT%201"

Questions

Common ClickHouse questions

Native TCP or HTTP endpoint?+

Both. The dashboard exposes the native protocol for the official clients (clickhouse-client, JDBC) and the HTTP endpoint for tools and runtimes that prefer it.

Will my existing ClickHouse driver work?+

Yes. Standard drivers (clickhouse-driver, ch-go, JDBC, Python clickhouse-client) connect over the published endpoint without changes.

Provisioning time?+

ClickHouse takes about 1 to 2 minutes to provision on the first create because of its larger binary, then hibernation wake is a few seconds. Subsequent restarts are fast.

Can I run ClickHouse locally?+

Yes. SpinDB manages local ClickHouse versions with the same one-command install; Layerbase Desktop adds the dashboard on top.

How does pricing work?+

ClickHouse is a paid-tier engine on Cloud. The cost preview in the create flow shows the monthly cost before you provision; the Free plan covers other engines for prototyping.