Update app.py

app.py

@@ -3,6 +3,7 @@ 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
@@ -12,6 +13,7 @@ import logging
 import g4f
 from g4f.client import Client
 
+
 # =====================================================
 # LOGGING
 # =====================================================
@@ -19,21 +21,30 @@ from g4f.client import Client
 logging.basicConfig(level=logging.INFO)
 logger = logging.getLogger(__name__)
 
+
 # =====================================================
 # CONFIG
 # =====================================================
 
 API_KEY = "sk-your-secret-key"
 
+# timeout لمنع التعليق الأبدي
+REQUEST_TIMEOUT = 45
+
+# retry خفيف
+MAX_RETRIES = 2
+
+
 # =====================================================
 # FASTAPI
 # =====================================================
 
 app = FastAPI(
     title="Universal AI Gateway",
-    version="6.0.0"
+    version="4.2.0"
 )
 
+
 # =====================================================
 # CORS
 # =====================================================
@@ -46,6 +57,7 @@ app.add_middleware(
     allow_headers=["*"],
 )
 
+
 # =====================================================
 # MODELS
 # =====================================================
@@ -60,7 +72,8 @@ class ChatRequest(BaseModel):
     messages: List[Message]
     stream: bool = False
     temperature: Optional[float] = 0.7
-    max_tokens: Optional[int] = 4096
+    max_tokens: Optional[int] = 32000
+
 
 # =====================================================
 # AUTH
@@ -90,6 +103,7 @@ def verify_api_key(req: Request):
 
     return True
 
+
 # =====================================================
 # ROOT
 # =====================================================
@@ -100,9 +114,10 @@ async def root():
     return {
         "status": "online",
         "service": "Universal AI Gateway",
-        "version": "6.0.0"
+        "version": "4.2.0"
     }
 
+
 # =====================================================
 # MODELS
 # =====================================================
@@ -111,20 +126,62 @@ async def root():
 async def get_models():
 
     models_data = []
-    existing_models = set()
+
+    # =================================================
+    # MODELS THAT WORK WELL
+    # =================================================
+
+    fallback_models = [
+
+        # GPT
+        "gpt-4o-mini",
+        "gpt-4o",
+        "gpt-4",
+        "gpt-3.5-turbo",
+
+        # Claude
+        "claude-3-haiku",
+
+        # Llama
+        "llama-3.1-70b",
+
+        # Mixtral
+        "mixtral-8x7b",
+
+        # Deepseek
+        "deepseek-chat",
+
+        # Gemini
+        "gemini-pro",
+
+        # =================================================
+        # COHERE FAMILY
+        # =================================================
+
+        "command-r",
+        "command-r-plus",
+        "command-r7b",
+        "command",
+        "command-nightly",
+
+        # Additional Cohere-style names
+        "cohere-command-r",
+        "cohere-command-r-plus",
+    ]
+
+    added_models = set()
 
     try:
 
-        # ترك g4f يسترد النماذج تلقائياً بدون تقييد
         if hasattr(g4f.models, "_all_models"):
 
             all_models = list(g4f.models._all_models)
 
-            for model in all_models:
+            for model in all_models[:100]:
 
                 model_name = str(model)
 
-                if model_name not in existing_models:
+                if model_name not in added_models:
 
                     models_data.append({
                         "id": model_name,
@@ -133,18 +190,92 @@ async def get_models():
                         "owned_by": "g4f"
                     })
 
-                    existing_models.add(model_name)
+                    added_models.add(model_name)
 
     except Exception as e:
 
         logger.error(f"Models error: {e}")
 
+    # fallback models
+    for model in fallback_models:
+
+        if model not in added_models:
+
+            models_data.append({
+                "id": model,
+                "object": "model",
+                "created": int(time.time()),
+                "owned_by": "g4f"
+            })
+
+            added_models.add(model)
+
     return {
         "object": "list",
-        "data": models_data,
-        "total": len(models_data)
+        "data": 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()
+
+            # timeout لمنع التعليق الأبدي
+            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
 # =====================================================
@@ -179,16 +310,7 @@ async def chat_completions(
 
             try:
 
-                # لا forcing
-                # لا providers يدوية
-                # لا routing
-                # لا watchdog
-                # لا timeout
-                # نترك g4f يعمل بطبيعته
-
-                client = Client()
-
-                response = client.chat.completions.create(
+                response = await safe_completion(
                     model=body.model,
                     messages=messages,
                     stream=True
@@ -202,16 +324,11 @@ async def chat_completions(
 
                     try:
 
-                        # تجاهل image responses
-                        if hasattr(chunk, "images"):
-                            continue
-
                         content = ""
 
                         if (
                             hasattr(chunk, "choices")
                             and chunk.choices
-                            and len(chunk.choices) > 0
                             and chunk.choices[0].delta
                             and chunk.choices[0].delta.content
                         ):
@@ -250,10 +367,19 @@ async def chat_completions(
                             f"Chunk error: {chunk_error}"
                         )
 
+                # provider فتح stream بدون محتوى
                 if not has_content:
 
-                    logger.warning(
-                        f"Empty stream | model={body.model}"
+                    error_payload = {
+                        "error": {
+                            "message": "Provider returned empty stream",
+                            "type": "empty_stream"
+                        }
+                    }
+
+                    yield (
+                        f"data: "
+                        f"{json.dumps(error_payload)}\n\n"
                     )
 
                 final_payload = {
@@ -309,12 +435,10 @@ async def chat_completions(
 
     try:
 
-        client = Client()
-
-        response = await asyncio.to_thread(
-            client.chat.completions.create,
+        response = await safe_completion(
             model=body.model,
-            messages=messages
+            messages=messages,
+            stream=False
         )
 
         assistant_message = ""
@@ -360,6 +484,7 @@ async def chat_completions(
             detail=str(e)
         )
 
+
 # =====================================================
 # RUN
 # =====================================================