butterbase-ai / butterbase-oss

What is butterbase-oss?

Butterbase is an open-source backend-as-a-service built in TypeScript that gives you Postgres, authentication, file storage, serverless functions, an AI gateway, and a built-in Model Context Protocol server. It runs on Docker and is designed for teams that want Firebase-like convenience without vendor lock-in. The stack uses Fastify for the API layer, Deno for executing serverless functions, and Postgres with row-level security for data isolation.

Why is it gaining traction?

The AI gateway and MCP server are the differentiators. Rather than wiring up multiple AI providers manually, you route all LLM traffic through Butterbase and get unified billing, logging, and key management in one place. The MCP server lets AI agents interact with your app directly, which is increasingly relevant as agentic workflows become mainstream. Self-hosting is a first-class concern: the Docker setup gives you a complete local environment with no cloud dependencies, and the billing system runs as a no-op when self-hosted.

Who should use this?

Teams building AI-powered applications who want to avoid stitching together separate auth, database, storage, and AI services. Developers tired of vendor pricing who need a self-hosted path forward. Hackathon teams that want a full backend stack running locally in minutes. If you need multi-region orchestration or want someone else to manage infrastructure, the managed butterbase.ai offering exists, but the open-source version handles single-region deployments well.

Verdict

Butterbase is early-stage (77 stars, v0.1.0) with a credibility score of 0.85%, so production use requires careful evaluation. The documentation is functional but the setup involves multiple manual steps including running migrations separately from container startup. That said, the architecture is sound and the feature set is competitive with commercial alternatives. Worth watching if you need an open-source BaaS with AI-first design.