Tinybird

• CategoryOpen-source / self-host
• PricingFreemium
• Starts atFree / $25
• LicenseProprietary
• Created2020
• GitHub starsclosed
• Script sizeBYO
• CookielessNo
• Self-hostNo

Tinybird is not a finished analytics product, it is the infrastructure layer for building your own. Hosted ClickHouse with a SQL-to-API workflow. Used by teams that need to ship user-facing analytics dashboards inside their own product (e.g. analytics-as-a-feature for SaaS). Generous free tier; usage-based pricing scales sensibly.

• SaaS companies shipping user-facing analytics as a product feature
• Teams that have outgrown PostHog / Mixpanel and need custom data models
• Engineers who want ClickHouse without operating ClickHouse

Skip Tinybird if you want a finished dashboard product, it is BYO frontend. Skip if you do not have engineering capacity to write SQL and build dashboards.

Tinybird is the right pick when analytics is a product feature, not a back-office tool. For internal analytics, PostHog wins on time-to-value. For analytics-shipped-to-customers, Tinybird is the modern answer.

ClickHouse is the right database for analytical queries over very large tables and the wrong thing to operate when data infrastructure is not your team's job. Tinybird's proposition is to keep the query engine and remove the cluster: you stream events in over HTTP or from Kafka, define pipes as SQL, and each pipe publishes as a parameterised, authenticated REST endpoint your front end calls directly. There is no serving layer to write and no ingestion service to maintain, which is usually a few sprints you were about to spend. Query latency is the reason people stay, because fast aggregations over huge tables are what make an in-product analytics tab feel instant rather than apologetic. The workflow is version-controllable and fits normal CI, which matters because these pipes become production API surface, not analyst scratch work.

Be clear about the scope. Tinybird gives you ingestion, storage, query, and an API. It does not give you a tracker script, a session model, bot filtering, a dashboard, charts, or any of the product decisions PostHog and Mixpanel have already made on your behalf. You are writing the SQL, designing the schema, choosing aggregation windows, and building the interface, and you need somebody comfortable with ClickHouse's particular way of thinking about sorting keys and materialised views. That is entirely reasonable when the output is a customer-facing feature you will charge for. It is a poor use of a team's time when the output is a chart your marketing lead checks weekly. The honest test: if nobody outside your company will ever see these numbers, buy a finished product instead.

• Official homepage →

The 30-min call is where your analytics choice becomes a real implementation, an event-tracking plan that fits your product, and a privacy story that does not need a 4-screen cookie banner.