video1.0<iframe src="https://www.loom.com/embed/6cb57484c5444c9aa0585db1a1b17bb5" frameborder="0" width="1662" height="1246" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>12461662Loomhttps://www.loom.com12461662https://cdn.loom.com/sessions/thumbnails/6cb57484c5444c9aa0585db1a1b17bb5-c39aa0c1c4f3b5e2.gif97.792This Loom demonstrates using Headroom as a guardrail in LightLLM. It shows creating a Headroom guardrail by adding the name and impression and selecting the provider as Headroom, including using a self-hosted API base and creating the guardrail key, then attaching it to a key. The video explains selecting the guardrail in the guardrails options and calling the chat completions or messages endpoint with the virtual key, provider name, and guardrail key. The logs confirm that the guardrail is evaluated and headroom compression is applied.