add grok 4

README.md

@@ -45,6 +45,7 @@ pip install -r requirements.txt
 export HF_TOKEN="your_huggingface_token"
 export TAVILY_API_KEY="your_tavily_api_key"  # Optional, for web search feature
 export DASHSCOPE_API_KEY="your_dashscope_api_key"  # Required for Qwen3-30B models via DashScope
+export POE_API_KEY="your_poe_api_key"  # Required for GPT-5 and Grok-4 via Poe
 ```
 
 ## Usage
@@ -79,6 +80,8 @@ python app.py
 - Qwen3-30B-A3B-Thinking-2507 (via DashScope)
 - SmolLM3-3B
 - GLM-4.1V-9B-Thinking (multimodal)
+- GPT-5 (via Poe)
+- Grok-4 (via Poe)
 
 ## Input Options
 

app.py

@@ -495,6 +495,11 @@ AVAILABLE_MODELS = [
         "name": "GPT-5",
         "id": "gpt-5",
         "description": "OpenAI GPT-5 model for advanced code generation and general tasks"
+    },
+    {
+        "name": "Grok-4",
+        "id": "grok-4",
+        "description": "Grok-4 model via Poe (OpenAI-compatible) for advanced tasks"
     }
 ]
 
@@ -602,6 +607,12 @@ def get_inference_client(model_id, provider="auto"):
             api_key=os.getenv("POE_API_KEY"),
             base_url="https://api.poe.com/v1"
         )
+    elif model_id == "grok-4":
+        # Use Poe (OpenAI-compatible) client for Grok-4 model
+        return OpenAI(
+            api_key=os.getenv("POE_API_KEY"),
+            base_url="https://api.poe.com/v1"
+        )
     elif model_id == "step-3":
         # Use StepFun API client for Step-3 model
         return OpenAI(
@@ -2554,6 +2565,13 @@ This will help me create a better design for you."""
                     stream=True,
                     max_tokens=16384
                 )
+            elif _current_model["id"] == "grok-4":
+                completion = client.chat.completions.create(
+                    model="Grok-4",
+                    messages=messages,
+                    stream=True,
+                    max_tokens=16384
+                )
             else:
                 completion = client.chat.completions.create(
                     model=_current_model["id"],