Deploy Flowise + Bifrost | Visual AI Agent Builder with a Built In Gateway
Flowise + Bifrost | Visual AI Agent Builder with a Built In Gateway
Just deployed
/var/lib/postgresql/data
Bifrost
Just deployed
/app/data
Flowise
Just deployed
/root/.flowise
Deploy and Host Flowise with Bifrost on Railway
Building agents and flows usually means pasting a provider key into every model node, with no failover, no shared budget, and no single place to see what the AI is costing. This template fixes that. Flowise gives you the visual drag and drop builder, and Bifrost sits underneath as the gateway every model node routes through, so one connection reaches OpenAI, Anthropic, Google, and more, with provider failover and request logging built in.
What you get
Flowise is an open source visual builder for LLM agents, RAG pipelines, and chat flows. Bifrost is an open source AI gateway that routes to 20+ providers behind one OpenAI compatible API. You point a Flowise model node at Bifrost once, and from then on your flows reach every provider through it, so you manage providers and keys in Bifrost instead of inside each flow.
About Hosting This Stack
The stack runs three services: Flowise as the builder with its own embedded storage on a volume, Bifrost as the AI gateway, and Postgres as Bifrost storage for configuration and request logs. After deploy, add a provider key in Bifrost, then in Flowise add a ChatOpenAI node and set its base URL to Bifrost over the private network. Every model call then flows through Bifrost with failover and full logging.
Common Use Cases
- Visual AI agents and RAG pipelines that can switch providers without editing every flow
- Chatbots and support assistants with one place to control AI cost and keys
- A shared builder for a team, with all model traffic logged through one gateway
Dependencies for This Stack
- Postgres for Bifrost configuration and request log storage
Deployment Dependencies
- Flowise: https://github.com/FlowiseAI/Flowise
- Bifrost docs: https://docs.getbifrost.ai
- Bifrost Docker Hub: https://hub.docker.com/r/maximhq/bifrost
Why Deploy This Stack on Railway?
Railway runs your whole stack on one platform, so you do not wire services together by hand and can scale each piece as load grows. By deploying Flowise with Bifrost on Railway, you get a private visual AI builder connected to every major provider in minutes, with cost control and logging built in.
Template Content
Bifrost
maximhq/bifrost:latestFlowise
flowiseai/flowise
