diff --git a/api-reference/assistant/create-assistant-message.mdx b/api-reference/assistant/create-assistant-message.mdx
index b2dd38479..83e5784af 100644
--- a/api-reference/assistant/create-assistant-message.mdx
+++ b/api-reference/assistant/create-assistant-message.mdx
@@ -2,6 +2,69 @@
 openapi: POST /assistant/{domain}/message
 ---
 
+The AI Discovery Assistant API provides intelligent, context-aware responses based on your documentation. The API returns responses as a **streaming text response**, allowing you to display answers progressively as they're generated.
+
+## Streaming response
+
+This endpoint returns a **streaming response** using Server-Sent Events (SSE). The response body contains a readable stream that you must process chunk by chunk to receive the complete answer.
+
+### Complete example
+
+Here's a complete example showing how to send a message and handle the streaming response:
+
+```javascript
+const domain = process.env.MINTLIFY_DOMAIN!;
+const apiKey = process.env.MINTLIFY_API_KEY!;
+const userMessage = process.argv.slice(2).join(' ') || 'How do I get started?';
+
+const response = await fetch(`https://api-dsc.mintlify.com/v1/assistant/${domain}/message`, {
+  method: 'POST',
+  headers: {
+    'Authorization': `Bearer ${apiKey}`,
+    'Content-Type': 'application/json',
+  },
+  body: JSON.stringify({
+    fp: 'user-fingerprint-' + Date.now(),
+    threadId: 'my-thread-id',
+    messages: [{
+      id: 'msg-' + Date.now(),
+      role: 'user',
+      content: userMessage,
+      parts: [{ type: 'text', text: userMessage }]
+    }],
+    retrievalPageSize: 5,
+    filter: null, // { version: "string", language: "string" }
+  }),
+});
+
+// Handle the streaming response
+const reader = response.body!.getReader();
+const decoder = new TextDecoder();
+
+while (true) {
+  const { done, value } = await reader.read();
+  if (done) break;
+  process.stdout.write(decoder.decode(value, { stream: true }));
+}
+```
+
+### Key parameters
+
+- **`fp`** - A unique fingerprint to identify the user session
+- **`threadId`** - Identifier for the conversation thread to maintain context
+- **`messages`** - Array of message objects with `id`, `role`, `content`, and `parts`
+- **`retrievalPageSize`** - Number of documentation chunks to retrieve for context (default: 5)
+- **`filter`** - Optional filter to limit search to specific versions or languages
+
+### Handling the stream
+
+The response is a readable stream that must be processed using the Streams API:
+
+1. Get a reader from `response.body.getReader()`
+2. Create a `TextDecoder` to convert bytes to text
+3. Read chunks in a loop until `done` is true
+4. Decode each chunk and process the text
+
 ## Rate limits
 
 The assistant API has the following limits:
@@ -10,8 +73,19 @@ The assistant API has the following limits:
 - 10,000 requests per Mintlify organization per hour
 - 10,000 requests per IP per day
 
-## Suggested usage
+## Frontend integration
+
+For React applications, use the [useChat hook from ai-sdk](https://ai-sdk.dev/docs/reference/ai-sdk-ui/use-chat#usechat) to send requests and handle streaming responses automatically.
 
-For best results, use the [useChat hook from ai-sdk](https://ai-sdk.dev/docs/reference/ai-sdk-ui/use-chat#usechat) to send requests and handle responses.
+You can set `fp`, `threadId`, and `filter` in the `body` field of the options parameter passed to the hook:
 
-You can set `fp`, `threadId`, and `filter` in the `body` field of the options parameter passed to the hook.
\ No newline at end of file
+```javascript
+const { messages, input, handleInputChange, handleSubmit } = useChat({
+  api: `/api/chat`,
+  body: {
+    fp: 'user-fingerprint-123',
+    threadId: 'conversation-thread-1',
+    filter: { version: 'v2', language: 'javascript' }
+  }
+});
+```
\ No newline at end of file