Open-source backend platform that gives coding agents a database, auth, storage, functions, hosting and an AI model gateway.
OtterMind/Nubase
Open-source, self-hostable AI-native backend that gives AI-generated apps a database, auth, storage, memory, functions and CDN hosting in one service.
Open-source alternative to:Supabase
- Stars
- 620
- Forks
- 73
- Open issues
- 5
- Last commit
- 19 hours ago
What is it?
Nubase is a backend and deploy layer intended to be driven by coding agents such as Claude Code or Codex through MCP tools. It bundles eight modules in one self-hostable service: Database, Auth, Storage, Assets, Functions, AI Gateway, Memory, and scheduled jobs (cron). Each project gets its own PostgreSQL database with a PostgREST-compatible REST API, and Auth follows Supabase-style conventions including JWTs and Row Level Security.
The project targets the gap between an AI-generated prototype and a running app. An agent can model tables, deploy edge functions, publish a frontend to a public CDN path, and schedule recurring work. Memory is a built-in primitive with LLM-based fact extraction and hybrid retrieval over pgvector and full-text search, rather than a separate vector store script.
The all-in-one Docker image bundles PostgreSQL, Redis, the Java backend and the Studio dashboard, and one control plane can provision multiple isolated project databases. The README lists Realtime and backups/HA as not yet implemented.
Key features
- 01Per-project PostgreSQL database with a PostgREST-compatible REST API and Row Level Security.
- 02Supabase-style auth with JWTs, OAuth providers, MFA and per-project API key roles.
- 03S3-compatible object storage with public and private buckets and signed URLs.
- 04Edge functions and static asset publishing that agents can deploy over MCP.
- 05Built-in memory API with LLM fact extraction and hybrid pgvector and full-text retrieval.
- 06One-line Docker image that bundles PostgreSQL, Redis, backend and Studio.
Use cases
- Turning a prototype generated by Claude Code or Codex into a live app with a database and login.
- Giving AI assistants persistent per-user memory backed by your own Postgres.
- Self-hosting a multi-project backend where each project has an isolated database.
- Publishing a generated frontend with edge functions and scheduled jobs on one self-hosted platform.
Quick start
- 1
Install the Nubase skills and MCP config for Claude Code or Codex
npx -y nubase_cli@latest install-skills - 2
Run your own Nubase instance with the all-in-one Docker image
docker run -d --name nubase -p 9999:9999 -p 5432:5432 -v nubase_data:/data <your-namespace>/nubase:latest - 3
Open Studio, create a project, and click Provision to initialize its database
Things to know
- The README describes Nubase as early-stage.
- Realtime is not implemented yet.
- Review the admin and management endpoints before exposing a server to the public internet.
In collections
Topics
FAQ
- What is Nubase?
- Open-source, self-hostable AI-native backend that gives AI-generated apps a database, auth, storage, memory, functions and CDN hosting in one service. Nubase is a backend and deploy layer intended to be driven by coding agents such as Claude Code or Codex through MCP tools. It bundles eight modules in one self-hostable service: Database, Auth, Storage, Assets, Functions, AI Gateway, Memory, and scheduled jobs (cron). Each project gets its own PostgreSQL database with a PostgREST-compatible REST API, and Auth follows Supabase-style conventions including JWTs and Row Level Security.
- Is Nubase free to use? What license does it have?
- Nubase is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.
- Is Nubase still maintained?
- Actively maintained. The last commit to Nubase was on Nov 7, 2026.
- What can I use Nubase for?
- Turning a prototype generated by Claude Code or Codex into a live app with a database and login. Giving AI assistants persistent per-user memory backed by your own Postgres. Self-hosting a multi-project backend where each project has an isolated database. Publishing a generated frontend with edge functions and scheduled jobs on one self-hosted platform.
- How do I install Nubase?
- Install the Nubase skills and MCP config for Claude Code or Codex: npx -y nubase_cli@latest install-skills Run your own Nubase instance with the all-in-one Docker image: docker run -d --name nubase -p 9999:9999 -p 5432:5432 -v nubase_data:/data <your-namespace>/nubase:latest Open Studio, create a project, and click Provision to initialize its database
- Where can I get Nubase?
- The source code and installation instructions are on GitHub: https://github.com/OtterMind/Nubase
Summary generated by AI from the project's README. GitHub stats refresh daily.
Related projects
A local-first Markdown memory workspace that lets AI agents store, search, and consolidate long-term knowledge as editable files.
A token-efficient, microkernel AI agent for CLI, Web UI and chat channels that routes each turn to the cheapest capable model.
Local, evidence-first memory runtime for AI agents that filters writes, tracks fact changes over time, and explains each recalled result.
A local-first memory and decision layer for AI agents that stores context as Markdown and JSONL files you own, usable across several IDEs and LLMs.
A demo customer service agent for a fictional airline, built with LangGraph and RAG to query Google Cloud databases through MCP Toolbox.