Update app.py

app.py

@@ -5,23 +5,39 @@ import io
 import json
 import requests
 
-HF_VLM_API = "https://api-inference.huggingface.co/models/Qwen/Qwen2-VL-7B-Instruct"
-HF_TOKEN = "ROBOT_MCP_TOKEN"  # API TOKEN
+OPENROUTER_URL = "https://openrouter.ai/api/v1/chat/completions"
+OPENROUTER_KEY = "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"  # OpenRouter key
 
 def call_vlm_api(img: Image):
     # encode image to bytes
     buf = io.BytesIO()
     img.save(buf, format="JPEG")
     img_bytes = buf.getvalue()
+    img_b64 = base64.b64encode(img_bytes).decode("utf-8")
 
-    headers = {"Authorization": f"Bearer {HF_TOKEN}"}
-    payload = {"inputs": [{"image": img_bytes, "text": "Describe the image in detail."}]}
-    
-    resp = requests.post(HF_VLM_API, headers=headers, json=payload, timeout=60)
+    headers = {"Authorization": f"Bearer {OPENROUTER_KEY}"}
+    payload = {
+        "model": "qwen/qwen2.5-vl-32b-instruct:free",
+        "messages": [
+            {
+                "role": "user",
+                "content": [
+                    {"type": "input_text", "text": "Describe the image in detail."},
+                    {"type": "input_image", "image_base64": img_b64}
+                ]
+            }
+        ],
+        "temperature": 0.2
+    }
+
+    resp = requests.post(OPENROUTER_URL, headers=headers, json=payload, timeout=60)
     if resp.status_code == 200:
-        return resp.json()[0].get("generated_text", "")
+        try:
+            return resp.json()["choices"][0]["message"]["content"][0]["text"]
+        except Exception:
+            return "Error parsing VLM response"
     else:
-        return f"VLM API error: {resp.status_code}"
+        return f"OpenRouter API error: {resp.status_code}"
 
 def process(payload: dict):
     try: