Skip to main content
These components help you serve, build, trace, and evaluate LLM applications. LiteLLM, Open WebUI, Langflow, and Langfuse can also run as separate project instances.

LiteLLM

LiteLLM is an OpenAI-compatible gateway in front of the model providers and self-hosted models you use. Applications call LiteLLM instead of calling each provider directly, and LiteLLM handles keys, routing, and usage tracking.

Open WebUI

Open WebUI is a chat interface for the models available through LiteLLM.
  • URL: https://open-webui.<base-hostname>

Langflow

Langflow is a visual builder for LLM flows and agents. Langflow can call LiteLLM inside the cluster.

Langfuse

Langfuse records traces, evaluations, and prompt versions for LLM applications.
  • URL: https://langfuse.<base-hostname>

Phoenix

Phoenix is an observability and evaluation tool for LLM applications.
  • URL: https://phoenix.<base-hostname>

Promptfoo

Promptfoo tests and compares prompts and models.
  • URL: https://promptfoo.<base-hostname>