From 8244ad1f0e5568fefda9467dc3e77db28d0e94c5 Mon Sep 17 00:00:00 2001 From: Ryan Crabbe Date: Mon, 23 Feb 2026 09:50:02 -0800 Subject: [PATCH] docs: tweak benchmarks wording --- docs/my-website/docs/benchmarks.md | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/docs/my-website/docs/benchmarks.md b/docs/my-website/docs/benchmarks.md index 47480d3dc63..5ed2263d05b 100644 --- a/docs/my-website/docs/benchmarks.md +++ b/docs/my-website/docs/benchmarks.md @@ -7,7 +7,7 @@ Benchmarks for LiteLLM Gateway (Proxy Server) tested against a fake OpenAI endpo ## Setting Up Benchmarking with Network Mock -The fastest way to benchmark proxy overhead is using `network_mock` mode. This intercepts outbound requests at the httpx transport layer and returns canned responses — no external endpoint needed. +The fastest way to benchmark proxy overhead is using `network_mock` mode. This intercepts outbound requests at the httpx transport layer and returns canned responses, no need for setting up a mock provider. **1. Create a proxy config:** @@ -41,7 +41,7 @@ litellm --config benchmark_config.yaml --port 4000 --num_workers 8 python scripts/benchmark_mock.py --requests 2000 --max-concurrent 200 --runs 3 ``` -This measures pure proxy overhead (auth, routing, logging) without any network latency to a real or fake provider. +This measures pure proxy overhead on the hot path without any network latency to a real or fake provider. ## Setting Up a Fake OpenAI Endpoint