From cf580a499511ad155f6651d804f248872ab92b95 Mon Sep 17 00:00:00 2001
From: ishaan-jaff <ishaanjaffer0324@gmail.com>
Date: Tue, 10 Oct 2023 10:19:15 -0700
Subject: [PATCH] (docs) add cost tracking to proxy server

---
 docs/my-website/docs/proxy_server.md | 17 ++++++++++++++++-
 1 file changed, 16 insertions(+), 1 deletion(-)

diff --git a/docs/my-website/docs/proxy_server.md b/docs/my-website/docs/proxy_server.md
index 1f4f66f4a86..44e2f280714 100644
--- a/docs/my-website/docs/proxy_server.md
+++ b/docs/my-website/docs/proxy_server.md
@@ -248,7 +248,22 @@ response = openai.ChatCompletion.create(model="ollama/llama2", messages=[{"role"
 
 print(response)
 ```
-    
+
+### Tracking costs
+By default litellm proxy writes cost logs to litellm/proxy/cost.log
+```shell
+2023-10-09 14:48:18 - Model gpt-3.5-turbo-0613 Cost: 0.000047000
+2023-10-09 14:48:18 - Model gpt-3.5-turbo-0613 Cost: 0.000128700
+2023-10-09 14:48:20 - Model gpt-3.5-turbo-0613 Cost: 0.000523011
+2023-10-09 14:48:28 - Model ollama/llama2 Cost: 0.00004700
+2023-10-09 14:48:35 - Model ollama/code-llama Cost: 0.00004700
+2023-10-09 14:48:49 - Model azure/chatgpt-2 Cost: 0.000223011
+```
+
+You can view costs on the cli using 
+``shell
+litellm --cost
+```
 
 ### Ollama Logs
 Ollama calls can sometimes fail (out-of-memory errors, etc.).