Rawbbit
rawbbit.one
- Free plan
- Web
- API
- Self-hosted
What Rawbbit is
Rawbbit is an open-source game analytics pipeline for studios that want to control their data. It accepts batches of game events over HTTP and keeps the complete event detail without requiring a predefined schema. Dashboards cover retention, revenue, level progression, funnels, and cohorts, while SQL and AI agents provide additional ways to query the data. A read-only MCP server works with Codex, OpenCode, OpenClaw, Claude, and other MCP clients. Events move through NATS JetStream, partitioned Parquet files in S3-compatible storage, and ClickHouse. Data remains in infrastructure controlled by the customer unless Rawbbit is asked to host it. The JavaScript SDK is available; Unity and Unreal SDKs are in progress, with backend or plain HTTP alternatives. Rawbbit is free to self-serve at no cost, including the repository, deployment documentation, and core pipeline.
Who it suits
Rawbbit suits game studios seeking analytics they can own, especially teams wanting designers and producers to answer questions without a data engineer. It is less suitable for teams seeking zero-setup, plug-and-play dashboards.
Good to know, and what to watch
- Open-source Apache 2.0 license
- Dashboards cover retention, funnels, cohorts, and revenue
- Read-only MCP access for multiple AI clients
- Customer-controlled event infrastructure
- Free self-serve plan at no cost
- Not designed as a zero-setup dashboard
- Unity and Unreal SDKs are still in progress
- Self-hosting requires deployment work
Plans
Every plan and priceDone-for-you setup
Free self-serve
Free plan
The verdict
Rawbbit combines open event collection, ready-made game analytics, and read-only AI querying. Its free self-serve option and customer-controlled storage fit studios prepared to manage deployment rather than seeking a turnkey dashboard.
What the maker publishes
- AI integrations
- Its read-only MCP server supports Codex, OpenCode, OpenClaw, Claude, and other MCP clients for querying game analytics. source
- Analytics
- The stack includes ready-to-use dashboards for retention, revenue, level progression, funnels, and cohorts. source
- Data ownership
- Raw event data is stored in infrastructure the customer controls unless the customer asks Rawbbit to host it. source
- Data stack
- Events are buffered with NATS JetStream, written as partitioned Parquet to S3-compatible storage, and loaded into ClickHouse. source
- Event collection
- It accepts event batches over HTTP and stores the events in full detail without requiring a predefined schema. source
- Intended users
- Rawbbit targets game studios seeking analytics they can own, especially teams that want designers and producers to get answers without a data engineer. source
- License
- Rawbbit is released under the Apache 2.0 license, with source code and deployment materials public on GitHub. source
- Notable limitation
- The maker says Rawbbit is not a fit for teams seeking a zero-setup, plug-and-play dashboard product. source
- Pricing model
- The maker says there are no subscriptions, per-user fees, or per-event pricing; operating costs depend on the infrastructure used. source
- Product
- Rawbbit is an open-source game analytics pipeline that collects game events and serves them through dashboards, SQL, and AI agents. source
- SDK support
- A JavaScript SDK is available; native Unity and Unreal SDKs are in progress, and those games can send events through a backend or plain HTTP. source
- Security
- The MCP analytics interface is read-only, and optional code access is read-only and can be scoped to selected systems or level packs. source
- Support
- The maker offers setup and maintenance services and invites prospective users to book a demo or contact [email protected]. source
Checked against the maker's pages on 3 Oct 2026.