Langflow
Langflow is an open-source, low-code visual builder for constructing LLM-powered agents and RAG workflows by dragging and connecting components on a flow-based canvas. It is built on top of Python AI libraries and can export flows for use as APIs or embedded applications.
What gets deployed with it
- PostgreSQL HA Cluster — required Langflow persists flow definitions, execution history, and user credentials in PostgreSQL.
Baseline topologies
| Topology | Minimum nodes | Roles |
|---|---|---|
| single-node | 1 | langflow |
| ha-pair | 2 | langflow-api |
What drives the cost
- Concurrent flow executions
- Expected peak number of simultaneously executing AI flows or pipelines. Drives CPU and worker memory planning.
- Active flow builders
- Number of users concurrently editing or testing flows in the visual builder. Affects API server and UI resource needs.
Where it runs
bare_metal, dedicated, cloud_vm
Size Langflow and get a cost estimate