Skip to content
#

cliproxyapi

Here are 17 public repositories matching this topic...

A plausible-looking model-pool config in a popular OSS LLM proxy silently destroys prompt caching. yangble5 is the corrected config, a Claude-wire shim, and the benchmark that proves it: 99.53% token-weighted cache hit on warm rounds (cold round 0%), a 748,918-token prompt. One machine, one run, 2026-07-21. Not a model. Built on CLIProxyAPI.

  • Updated Aug 3, 2026
  • Python

Improve this page

Add a description, image, and links to the cliproxyapi topic page so that developers can more easily learn about it.

Curate this topic

Add this topic to your repo

To associate your repository with the cliproxyapi topic, visit your repo's landing page and select "manage topics."

Learn more