Update app.py

app.py

@@ -3,7 +3,6 @@ from fastapi.middleware.cors import CORSMiddleware
 from fastapi.responses import StreamingResponse, JSONResponse
 from pydantic import BaseModel
 from typing import List, Optional
-
 import asyncio
 import json
 import time
@@ -26,19 +25,13 @@ logger = logging.getLogger(__name__)
 
 API_KEY = "sk-your-secret-key"
 
-# timeout بسيط فقط لمنع التعليق الأبدي
-REQUEST_TIMEOUT = 45
-
-# retry خفيف
-MAX_RETRIES = 2
-
 # =====================================================
 # FASTAPI
 # =====================================================
 
 app = FastAPI(
     title="Universal AI Gateway",
-    version="5.0.0"
+    version="FINAL-5.1.0"
 )
 
 # =====================================================
@@ -77,7 +70,7 @@ def verify_api_key(req: Request):
 
     auth = req.headers.get("Authorization")
 
-    # السماح للاختبار
+    # السماح بدون مفتاح للاختبار
     if not auth:
         return True
 
@@ -107,7 +100,7 @@ async def root():
     return {
         "status": "online",
         "service": "Universal AI Gateway",
-        "version": "5.0.0"
+        "version": "FINAL-5.1.0"
     }
 
 # =====================================================
@@ -118,11 +111,7 @@ async def root():
 async def get_models():
 
     models_data = []
-    added_models = set()
-
-    # =================================================
-    # DYNAMIC MODELS FROM G4F
-    # =================================================
+    existing_models = set()
 
     try:
 
@@ -134,7 +123,7 @@ async def get_models():
 
                 model_name = str(model)
 
-                if model_name not in added_models:
+                if model_name not in existing_models:
 
                     models_data.append({
                         "id": model_name,
@@ -143,43 +132,43 @@ async def get_models():
                         "owned_by": "g4f"
                     })
 
-                    added_models.add(model_name)
+                    existing_models.add(model_name)
 
     except Exception as e:
 
         logger.error(f"Models error: {e}")
 
     # =================================================
-    # IMPORTANT WORKING MODELS
+    # IMPORTANT MODELS THAT ACTUALLY WORK
     # =================================================
 
-    important_models = [
+    extra_models = [
 
-        # Cohere family
+        # Cohere
         "command-r",
         "command-r-plus",
         "command-r7b",
         "command-a",
 
-        # Kimi
-        "kimi-k2",
-
         # DeepSeek
         "deepseek-chat",
         "deepseek-v3",
         "deepseek-r1",
 
-        # Gemini
-        "gemini-2.5-flash",
+        # Kimi
+        "kimi-k2",
 
         # GPT
         "gpt-4",
         "gpt-4o",
+
+        # Gemini
+        "gemini-2.5-flash",
     ]
 
-    for model in important_models:
+    for model in extra_models:
 
-        if model not in added_models:
+        if model not in existing_models:
 
             models_data.append({
                 "id": model,
@@ -188,7 +177,7 @@ async def get_models():
                 "owned_by": "g4f"
             })
 
-            added_models.add(model)
+            existing_models.add(model)
 
     return {
         "object": "list",
@@ -196,64 +185,6 @@ async def get_models():
         "total": len(models_data)
     }
 
-# =====================================================
-# SAFE COMPLETION
-# =====================================================
-
-async def safe_completion(
-    model,
-    messages,
-    stream=False
-):
-
-    last_error = None
-
-    for attempt in range(MAX_RETRIES):
-
-        try:
-
-            logger.info(
-                f"Attempt {attempt + 1} | model={model}"
-            )
-
-            client = Client()
-
-            response = await asyncio.wait_for(
-                asyncio.to_thread(
-                    client.chat.completions.create,
-                    model=model,
-                    messages=messages,
-                    stream=stream
-                ),
-                timeout=REQUEST_TIMEOUT
-            )
-
-            logger.info(
-                f"Success | model={model}"
-            )
-
-            return response
-
-        except asyncio.TimeoutError:
-
-            last_error = "Request timeout"
-
-            logger.warning(
-                f"Timeout | model={model}"
-            )
-
-        except Exception as e:
-
-            last_error = e
-
-            logger.warning(
-                f"Attempt failed {attempt + 1} | {e}"
-            )
-
-        await asyncio.sleep(1)
-
-    raise Exception(last_error)
-
 # =====================================================
 # CHAT COMPLETIONS
 # =====================================================
@@ -288,7 +219,16 @@ async def chat_completions(
 
             try:
 
-                response = await safe_completion(
+                client = Client()
+
+                # IMPORTANT:
+                # لا timeout
+                # لا retry wrappers
+                # لا asyncio.to_thread
+                # لا watchdog
+                # نترك g4f يعمل بطبيعته
+
+                response = client.chat.completions.create(
                     model=body.model,
                     messages=messages,
                     stream=True
@@ -302,12 +242,12 @@ async def chat_completions(
 
                     try:
 
-                        content = ""
-
                         # تجاهل image models
                         if hasattr(chunk, "images"):
                             continue
 
+                        content = ""
+
                         if (
                             hasattr(chunk, "choices")
                             and chunk.choices
@@ -352,16 +292,8 @@ async def chat_completions(
                 # stream فارغ
                 if not has_content:
 
-                    error_payload = {
-                        "error": {
-                            "message": "Empty stream",
-                            "type": "empty_stream"
-                        }
-                    }
-
-                    yield (
-                        f"data: "
-                        f"{json.dumps(error_payload)}\n\n"
+                    logger.warning(
+                        f"Empty stream | model={body.model}"
                     )
 
                 final_payload = {
@@ -417,10 +349,12 @@ async def chat_completions(
 
     try:
 
-        response = await safe_completion(
+        client = Client()
+
+        response = await asyncio.to_thread(
+            client.chat.completions.create,
             model=body.model,
-            messages=messages,
-            stream=False
+            messages=messages
         )
 
         assistant_message = ""