Update app.py

app.py

@@ -33,6 +33,17 @@ internals.LlamaModel.__del__ = safe_del
 
 
 
+##### final verdict
+
+# GLM 4.7 flash fast infrence
+#qwen 3 VL
+#mini max 2,5
+# qwen 3 coder next
+#gpt oss 120B
+#qwen 3 next  80b
+
+
+
 
 
 
@@ -96,12 +107,12 @@ huggingface_token = os.getenv("HUGGINGFACE_TOKEN")
 
 
 
-# hf_hub_download(
-#     repo_id="unsloth/Qwen3-Next-80B-A3B-Instruct-GGUF",
-#     filename="Qwen3-Next-80B-A3B-Instruct-Q4_K_M.gguf",
-#     local_dir="./models",
-#     token=huggingface_token
-# )
+hf_hub_download(
+    repo_id="unsloth/Qwen3-Next-80B-A3B-Instruct-GGUF",
+    filename="Qwen3-Next-80B-A3B-Instruct-Q4_K_M.gguf",
+    local_dir="./models",
+    token=huggingface_token
+)
 
 
 # hf_hub_download(
@@ -126,13 +137,17 @@ from huggingface_hub import snapshot_download
 #     token=huggingface_token      # only if gated/private
 # )
 
-snapshot_download(
-    repo_id="unsloth/gpt-oss-120b-GGUF",
-    repo_type="model",
-    local_dir="./models/",
-    allow_patterns=["Q8_0/*"],   # 👈 folder inside repo
-    token=huggingface_token      # only if gated/private
-)
+
+
+
+#### Deploy Minimax 2.5 insplace of  gpt oss 120b     its larger . and better and more recet leeases
+# snapshot_download(
+#     repo_id="unsloth/gpt-oss-120b-GGUF",
+#     repo_type="model",
+#     local_dir="./models/",
+#     allow_patterns=["Q8_0/*"],   # 👈 folder inside repo
+#     token=huggingface_token      # only if gated/private
+# )
 
 
 
@@ -302,7 +317,7 @@ demo = gr.ChatInterface(
                 "Qwen3-VL-32B-Thinking-Q8_0.gguf",
                 "Q8_0/gpt-oss-120b-Q8_0-00001-of-00002.gguf"
             ],
-            value="Q8_0/gpt-oss-120b-Q8_0-00001-of-00002.gguf",
+            value="Qwen3-Next-80B-A3B-Instruct-Q4_K_M.gguf",
             label="Model",
         ),
         gr.Textbox(