diff --git a/sources/agent/imageDescription.ts b/sources/agent/imageDescription.ts
index d13387e..1dd8837 100644
--- a/sources/agent/imageDescription.ts
+++ b/sources/agent/imageDescription.ts
@@ -1,5 +1,6 @@
 import { KnownModel, ollamaInference } from "../modules/ollama";
 import { groqRequest } from "../modules/groq-llama3";
+import { gptRequest } from "../modules/openai";
 
 
 export async function imageDescription(src: Uint8Array, model?: KnownModel): Promise<string> {
@@ -32,4 +33,22 @@ export async function llamaFind(question: string, images: string): Promise<strin
         , 
             question
     );
+}
+
+export async function openAIFind(question: string, images: string): Promise<string> {
+    return gptRequest(
+             `
+                You are a smart AI that need to read through description of a images and answer user's questions. 
+                
+                This are the provided images:
+                ${images}
+
+                DO NOT mention the images, scenes or descriptions in your answer, just answer the question.
+                DO NOT try to generalize or provide possible scenarios.
+                ONLY use the information in the description of the images to answer the question.
+                BE concise and specific.
+            `
+        , 
+            question
+    );
 }
\ No newline at end of file
diff --git a/sources/modules/openai.ts b/sources/modules/openai.ts
index d41482b..42debf4 100644
--- a/sources/modules/openai.ts
+++ b/sources/modules/openai.ts
@@ -71,13 +71,13 @@ export async function describeImage(imagePath: string) {
     }
 }
 
-export async function gptRequest(prompt: string) {
+export async function gptRequest(systemPrompt: string, userPrompt: string) {
     try {
         const response = await axios.post("https://api.openai.com/v1/chat/completions", {
-            model: "gpt-4",
+            model: "gpt-4o",
             messages: [
-                { role: "system", content: "You are a helpful assistant." },
-                { role: "user", content: prompt },
+                { role: "system", content: systemPrompt },
+                { role: "user", content: userPrompt },
             ],
         });
         return response.data;
@@ -88,4 +88,22 @@ export async function gptRequest(prompt: string) {
 }
 
 
-textToSpeech("Hello I am an agent")
\ No newline at end of file
+textToSpeech("Hello I am an agent")
+console.info(gptRequest(
+    `
+                You are a smart AI that need to read through description of a images and answer user's questions. 
+                
+                This are the provided images:
+                The image features a woman standing in an open space with a metal roof, possibly at a train station or another large building.
+                She is wearing a hat and appears to be looking up towards the sky.
+                The scene captures her attention as she gazes upwards, perhaps admiring something above her or simply enjoying the view from this elevated position.
+
+                DO NOT mention the images, scenes or descriptions in your answer, just answer the question.
+                DO NOT try to generalize or provide possible scenarios.
+                ONLY use the information in the description of the images to answer the question.
+                BE concise and specific.
+            `
+        , 
+            'where is the person?'
+
+))
\ No newline at end of file