Upload 13 files

README.md

@@ -35,22 +35,9 @@ curl -X POST https://Rox-Turbo-API.hf.space/chat \
   -d '{"messages":[{"role":"user","content":"Hello"}]}'
 ```
 
-### With Web Search (NEW!)
-
-Get real-time information from the web:
-
-```bash
-curl -X POST https://Rox-Turbo-API.hf.space/chat \
-  -H "Content-Type: application/json" \
-  -d '{
-    "messages":[{"role":"user","content":"Latest AI news?"}],
-    "websearch": true
-  }'
-```
-
 ## Usage
 
-### Basic Request
+### Request
 
 ```json
 {
@@ -62,18 +49,6 @@ curl -X POST https://Rox-Turbo-API.hf.space/chat \
 }
 ```
 
-### With Web Search
-
-```json
-{
-  "messages": [
-    {"role": "user", "content": "What's new in AI?"}
-  ],
-  "websearch": true,
-  "websearch_results": 5
-}
-```
-
 ### Response
 
 ```json
@@ -136,7 +111,6 @@ curl -X POST https://Rox-Turbo-API.hf.space/turbo45 \
 
 ## Documentation
 
-- [Web Search Guide](docs/WEBSEARCH.md) - Real-time web search
 - [Code Examples](docs/CODE.md) - Copy-paste code
 - [Developer Guide](docs/DEVELOPER_GUIDE.md) - Integration guide
 - [Models Guide](docs/MODELS.md) - Model details

python-version

@@ -0,0 +1 @@
+3.14

server.py

@@ -16,7 +16,6 @@ from pydantic import BaseModel, Field
 from openai import AsyncOpenAI
 import httpx
 import json
-from websearch import perform_web_search
 
 
 # Load environment variables
@@ -71,10 +70,10 @@ SYSTEM_PROMPT_MODE = os.getenv("SYSTEM_PROMPT_MODE", "full").strip().lower()
 
 # Model configurations
 ROX_CORE_MODEL = "z-ai/glm4.7"
-ROX_TURBO_MODEL = "deepseek-ai/deepseek-v3.2"
+ROX_TURBO_MODEL = "deepseek-ai/deepseek-v3.1-terminus"
 ROX_CODER_MODEL = "z-ai/glm5"
 ROX_TURBO_45_MODEL = "qwen/qwen3-coder-480b-a35b-instruct"
-ROX_ULTRA_MODEL = "deepseek-ai/deepseek-v4-pro"
+ROX_ULTRA_MODEL = "deepseek-ai/deepseek-v3.2"
 ROX_DYNO_MODEL = "moonshotai/kimi-k2.5"
 ROX_CODER_7_MODEL = "z-ai/glm-5.1"
 ROX_VISION_MODEL = "google/gemma-3-27b-it"
@@ -285,70 +284,6 @@ def _sse_headers() -> dict:
     }
 
 
-async def _handle_websearch(
-    app_: FastAPI,
-    messages: list,
-    websearch_enabled: bool,
-    num_results: int
-) -> list:
-    """
-    Handle web search if enabled and inject results into messages
-    
-    Args:
-        app_: FastAPI app instance
-        messages: Current message list
-        websearch_enabled: Whether web search is enabled
-        num_results: Number of search results to fetch
-    
-    Returns:
-        Updated messages list with search results if applicable
-    """
-    if not websearch_enabled or not messages:
-        return messages
-    
-    # Get the last user message as search query
-    last_user_msg = None
-    for msg in reversed(messages):
-        if msg.get("role") == "user":
-            last_user_msg = msg.get("content", "")
-            break
-    
-    if not last_user_msg:
-        return messages
-    
-    # Perform web search
-    http_client = getattr(app_.state, "http_client", None)
-    if not http_client:
-        logger.warning("HTTP client not available for web search")
-        return messages
-    
-    try:
-        success, search_results, error = await perform_web_search(
-            http_client,
-            last_user_msg,
-            num_results
-        )
-        
-        if success and search_results:
-            # Inject search results before the last user message
-            search_context = {
-                "role": "system",
-                "content": f"[Web Search Results]\n{search_results}\n\nUse these search results to provide an accurate, up-to-date answer to the user's question."
-            }
-            
-            # Insert search results before last user message
-            messages_copy = messages[:-1] + [search_context, messages[-1]]
-            logger.info("Web search completed: %d results", num_results)
-            return messages_copy
-        else:
-            if error:
-                logger.warning("Web search failed: %s", error)
-    except Exception as e:
-        logger.error("Web search error: %s", str(e))
-    
-    return messages
-
-
 # Helper function for streaming responses
 async def stream_response(
     app_: FastAPI,
@@ -476,8 +411,6 @@ class ChatRequest(BaseModel):
     top_p: Optional[float] = None
     max_tokens: Optional[int] = None
     stream: Optional[bool] = False
-    websearch: Optional[bool] = False
-    websearch_results: Optional[int] = 5
 
 
 class ChatResponse(BaseModel):
@@ -501,22 +434,12 @@ class HFResponseItem(BaseModel):
 
 @app.post("/chat")
 async def chat(req: ChatRequest):
-    """Rox Core - Main conversational model with streaming support and web search"""
+    """Rox Core - Main conversational model with streaming support"""
     messages: list = []
     system_prompt = _system_prompt_for("core")
     if system_prompt:
         messages.append({"role": "system", "content": system_prompt})
     messages.extend([m.model_dump() for m in req.messages])
-    
-    # Handle web search if enabled
-    if req.websearch:
-        messages = await _handle_websearch(
-            app,
-            messages,
-            req.websearch,
-            req.websearch_results or 5
-        )
-    
     temperature = _effective_temperature(req.temperature)
     top_p = _effective_top_p(req.top_p)
     max_tokens = _effective_max_tokens(req.max_tokens, 8192)
@@ -545,17 +468,12 @@ async def chat(req: ChatRequest):
 
 @app.post("/turbo")
 async def turbo(req: ChatRequest):
-    """Rox 2.1 Turbo - Fast and efficient with streaming and web search"""
+    """Rox 2.1 Turbo - Fast and efficient with streaming"""
     messages: list = []
     system_prompt = _system_prompt_for("turbo")
     if system_prompt:
         messages.append({"role": "system", "content": system_prompt})
     messages.extend([m.model_dump() for m in req.messages])
-    
-    # Handle web search if enabled
-    if req.websearch:
-        messages = await _handle_websearch(app, messages, req.websearch, req.websearch_results or 5)
-    
     temperature = _effective_temperature(req.temperature)
     top_p = _effective_top_p(req.top_p)
     max_tokens = _effective_max_tokens(req.max_tokens, 8192)
@@ -584,17 +502,12 @@ async def turbo(req: ChatRequest):
 
 @app.post("/coder")
 async def coder(req: ChatRequest):
-    """Rox 3.5 Coder - Specialized coding with streaming and web search"""
+    """Rox 3.5 Coder - Specialized coding with streaming"""
     messages: list = []
     system_prompt = _system_prompt_for("coder")
     if system_prompt:
         messages.append({"role": "system", "content": system_prompt})
     messages.extend([m.model_dump() for m in req.messages])
-    
-    # Handle web search if enabled
-    if req.websearch:
-        messages = await _handle_websearch(app, messages, req.websearch, req.websearch_results or 5)
-    
     temperature = _effective_temperature(req.temperature)
     top_p = _effective_top_p(req.top_p)
     max_tokens = _effective_max_tokens(req.max_tokens, 16384)
@@ -668,17 +581,12 @@ async def turbo45(req: ChatRequest):
 
 @app.post("/ultra")
 async def ultra(req: ChatRequest):
-    """Rox 5 Ultra - Most advanced with streaming and web search"""
+    """Rox 5 Ultra - Most advanced with streaming"""
     messages: list = []
     system_prompt = _system_prompt_for("ultra")
     if system_prompt:
         messages.append({"role": "system", "content": system_prompt})
     messages.extend([m.model_dump() for m in req.messages])
-    
-    # Handle web search if enabled
-    if req.websearch:
-        messages = await _handle_websearch(app, messages, req.websearch, req.websearch_results or 5)
-    
     temperature = _effective_temperature(req.temperature)
     top_p = _effective_top_p(req.top_p)
     max_tokens = _effective_max_tokens(req.max_tokens, 8192)