fix: OpenRouter not responding issue

.github/README.md

@@ -1,6 +1,6 @@
 # Axiom - A Docs Expert Agent
 
-Axiom is AI Agent specialized in modern AI frameworks, libraries and tools. It can assist in creating AI Agents, RAG systems, chatbots, authentication mechanisms, and even full-stack development. It is built with LangGraph, **MCP Docs Server**, Chainlit and OpenRouter Models, designed to help users create different projects using natural language instructions.
+Axiom is AI Agent specialized in modern AI frameworks, libraries and tools. It can assist in creating AI Agents, RAG systems, chatbots, authentication mechanisms, and even full-stack development. It is built with LangGraph, **MCP Docs Server**, Chainlit and Gemini Models, designed to help users create different projects using natural language instructions.
 
 ![AxiomAgent](/public/axiom.png)
 
@@ -8,7 +8,7 @@ Axiom is AI Agent specialized in modern AI frameworks, libraries and tools. It c
 
 - 🤖 Interactive chat interface
 - 📚 Access to multiple documentation sources
-- 🦾 Support for multiple models using OpenRouter unified API
+- 🦾 Support for multiple Gemini models
 - 🎨 Support for image processing and analysis
 - 📈 Use images and graphs to create production-ready code
 - 🛠️ Customizable model settings (temperature, model version)
@@ -39,7 +39,7 @@ The agent has access to following documentations:
 
 *   UV package manager
 *   Python 3.11+
-*   OpenRouter API Key
+*   Google Gemini API Key
 *   Docker (optional): If you intend to use the `Dockerfile`, you'll need Docker installed.
 
 ## Installation

chainlit.md

@@ -1,6 +1,6 @@
 # Axiom - A Docs Expert Agent
 
-Axiom is AI Agent specialized in modern AI frameworks, libraries and tools. It can assist in creating AI Agents, RAG systems, chatbots, authentication mechanisms, and even full-stack development. It is built with LangGraph, MCP Docs Server, Chainlit and OpenRouter Models, designed to help users create different projects using natural language instructions.
+Axiom is AI Agent specialized in modern AI frameworks, libraries and tools. It can assist in creating AI Agents, RAG systems, chatbots, authentication mechanisms, and even full-stack development. It is built with LangGraph, MCP Docs Server, Chainlit and Gemini Models, designed to help users create different projects using natural language instructions.
 
 ## Features
 
@@ -9,7 +9,7 @@ Axiom is AI Agent specialized in modern AI frameworks, libraries and tools. It c
     - **Agent Mode:** Ideal for complex tasks like brainstorming, code generation, and web apps creation.
     - **Chat Mode:** Suited for quick information retrieval and answering questions from the provided documentations.
 - 📚 Access to multiple documentation sources
-- 🦾 Support for multiple models using OpenRouter unified API
+- 🦾 Support for multiple Gemini models
 - 🎨 Support for image processing and analysis
 - 📈 Use images and graphs to create production-ready code
 - 🛠️ Customizable model settings (temperature, model version)

src/axiom/app.py

@@ -2,7 +2,6 @@
 import chainlit as cl
 from agent import make_graph
 
-from langchain_openai import ChatOpenAI
 from langchain_google_genai import ChatGoogleGenerativeAI
 from langchain_core.messages import AIMessageChunk, HumanMessage
 
@@ -105,12 +104,11 @@ async def on_chat_start():
                 id="model",
                 label="Gemini - Model",
                 values=[
-                    "deepseek/deepseek-chat-v3-0324", 
-                    "deepseek/deepseek-r1:free", 
-                    "google/gemini-2.0-flash:free", 
-                    "google/gemini-2.0-pro-exp-02-05:free", 
-                    "google/gemini-exp-1206:free", 
-                    "google/gemini-2.5-pro-exp-03-25:free"
+                    "gemini-2.0-flash", 
+                    "gemini-2.0-flash-lite",
+                    "gemini-2.0-pro-exp-02-05", 
+                    "gemini-exp-1206", 
+                    "gemini-2.5-pro-exp-03-25"
                     ],
                 initial_index=0,
             ),
@@ -126,10 +124,9 @@ async def on_chat_start():
     ).send()
 
     # Create model with given settings
-    model = ChatOpenAI(
+    model = ChatGoogleGenerativeAI(
         model=settings["model"],
-        api_key=os.getenv("OPENROUTER_API_KEY"),
-        base_url="https://openrouter.ai/api/v1",
+        api_key=os.getenv("GOOGLE_API_KEY"),
         temperature=settings["temperature"]
         )