Open-source AI infrastructure — self-hosted, API-first
Synaplan is fully open-source. Self-host it, connect via REST API, build custom plugins, or use it as the AI backbone for your own product — Apache 2.0, no lock-in.
Built for builders
100% open source
Fork, inspect, and contribute. Apache 2.0 licence — no lock-in, no black box. The full stack (PHP backend, Vue.js frontend, MariaDB, Qdrant) is on GitHub.
REST API
Every function exposed as an API endpoint. Build your own UI, automate workflows, or integrate Synaplan into existing tools — documented at docs.synaplan.com.
Plugin system
Extend Synaplan with purpose-built plugins for OpenCloud, WordPress, marketing, and more. Or publish your own under the metadist organisation.
Docker & self-hosting
Deploy in minutes with docker compose. Full stack ready out of the box — including local AI inference via Ollama so your data never leaves your infrastructure.
Multi-model routing
Route by cost, latency, or policy. Swap models without touching your integration code — GPT, Claude, Gemini, and any local model work interchangeably.
Vector search
Qdrant and MariaDB VECTOR for semantic RAG — no external embedding service required. Fine-tune retrieval per knowledge source and use case.
Up and running in 2 minutes
# Clone the repo
git clone https://github.com/metadist/synaplan
cd synaplan
# Start all services
docker compose up -d
# Open the app — Open http://localhost:5173 — full stack ready.
Extend Synaplan with plugins
Marketing, OpenCloud, WordPress, file management — and more on the way. All plugins are open-source and follow the same architecture.
Explore plugins →