diff --git a/docs/my-website/docs/proxy_api.md b/docs/my-website/docs/proxy_api.md
index 39234203e..29d054f27 100644
--- a/docs/my-website/docs/proxy_api.md
+++ b/docs/my-website/docs/proxy_api.md
@@ -1,33 +1,47 @@
 import TokenGen from '../src/components/TokenGen.js'
 
-# 🚨 LITELLM API (Access Claude-2/Llama2-70b/etc.)
+# 🚨 LiteLLM API (Access Claude-2,Llama2-70b,etc.)
 
 Use this if you're trying to add support for new LLMs and need access for testing: 
 
-Here's how to call it: 
+# usage
 
 ## Step 1: Save your LiteLLM API Key 
 
-This is your unique LiteLLM API Key. Save this for later use. 
+This is your unique LiteLLM API Key. It has a max budget of $100 which is reset monthly, and works across all models in [Supported Models](#supported-models). Save this for later use. 
 <TokenGen/>
 
 ## Step 2: Test a new LLM
 
-Now let's test if claude-2 is working in our code
+Now let's call **claude-2** (Anthropic) and **llama2-70b-32k** (TogetherAI).
+
 ```
 from litellm import completion 
 import os 
 
 # set env var
-os.environ["ANTHROPIC_API_KEY"] = "sk-litellm-1234" # 👈 replace with your unique key
+os.environ["ANTHROPIC_API_KEY"] = "sk-litellm-1234" # 👈 replace with your unique key from step 1
+os.environ["TOGETHERAI_API_KEY"] = "sk-litellm-1234" # 👈 replace with your unique key from step 1
 
 messages = [{"role": "user", "content": "Hey, how's it going?"}]
 
+# call claude
 response = completion(model="claude-2", messages=messages) 
 
+# call llama2-70b
+response = completion(model="togethercomputer/LLaMA-2-7B-32K", messages=messages) 
+
 print(response) 
 ```
 
+```
+### Testing Llama2-70b on TogetherAI 
+Let's call 
+
+```
+
+You can use this as a key for any of the [providers we support](./providers/)
+
 ## Supported Models
 
 * OpenAI models 
diff --git a/docs/my-website/src/components/TokenGen.js b/docs/my-website/src/components/TokenGen.js
index 1e89facc7..5ffa7d48a 100644
--- a/docs/my-website/src/components/TokenGen.js
+++ b/docs/my-website/src/components/TokenGen.js
@@ -1,5 +1,4 @@
 import React, { useState, useEffect } from 'react';
-import {v4 as uuidv4} from 'uuid'; 
 
 const CodeBlock = ({ token }) => {
   const codeWithToken = `${token}`;
@@ -15,20 +14,37 @@ const TokenGen = () => {
   const [token, setToken] = useState(null);
 
   useEffect(() => {
-    const generateToken = () => {
-      // Generate a special uuid/token "sk-litellm-<uuid>"
-      const newToken = `sk-litellm-${uuidv4()}`;
-      setToken(newToken);
-    };
+    const generateToken = async () => {
+      try {
+        const response = await fetch('https://proxy.litellm.ai/key/new', {
+          method: 'POST',
+          headers: {
+            'Content-Type': 'application/json',
+            'Authorization': 'Bearer sk-liteplayground',
+          },
+          body: JSON.stringify({'total_budget': 100})
+        });
+        
+        if (!response.ok) {
+          throw new Error('Network response was not ok');
+        }
+        
+        const data = await response.json();
 
-    generateToken();
-  }, []);
+        setToken(`${data.api_key}`);
+    } catch (error) {
+      console.error('Failed to fetch new token: ', error);
+    }
+  };
 
-  return (
-    <div>
-      <CodeBlock token={token} />
-    </div>
-  );
+  generateToken();
+}, []);
+
+return (
+  <div>
+    <CodeBlock token={token} />
+  </div>
+);
 };
 
-export default TokenGen;
\ No newline at end of file
+export default TokenGen;