Merge 15b9f90ae2 into fa4d01c23a

Implement backtesting strategy using TradingAgents with data verification and caching.

.env.enterprise.example

@@ -0,0 +1,5 @@
+# Azure OpenAI
+AZURE_OPENAI_API_KEY=
+AZURE_OPENAI_ENDPOINT=https://your-resource-name.openai.azure.com/
+AZURE_OPENAI_DEPLOYMENT_NAME=
+# OPENAI_API_VERSION=2024-10-21  # optional, required for non-v1 API

.env.example

@@ -3,4 +3,7 @@ OPENAI_API_KEY=
 GOOGLE_API_KEY=
 ANTHROPIC_API_KEY=
 XAI_API_KEY=
+DEEPSEEK_API_KEY=
+DASHSCOPE_API_KEY=
+ZHIPU_API_KEY=
 OPENROUTER_API_KEY=

README.md

@@ -140,10 +140,15 @@ export OPENAI_API_KEY=...          # OpenAI (GPT)
 export GOOGLE_API_KEY=...          # Google (Gemini)
 export ANTHROPIC_API_KEY=...       # Anthropic (Claude)
 export XAI_API_KEY=...             # xAI (Grok)
+export DEEPSEEK_API_KEY=...        # DeepSeek
+export DASHSCOPE_API_KEY=...       # Qwen (Alibaba DashScope)
+export ZHIPU_API_KEY=...           # GLM (Zhipu)
 export OPENROUTER_API_KEY=...      # OpenRouter
 export ALPHA_VANTAGE_API_KEY=...   # Alpha Vantage
 ```
 
+For enterprise providers (e.g. Azure OpenAI, AWS Bedrock), copy `.env.enterprise.example` to `.env.enterprise` and fill in your credentials.
+
 For local models, configure Ollama with `llm_provider: "ollama"` in your config.
 
 Alternatively, copy `.env.example` to `.env` and fill in your keys:

backtest

@@ -0,0 +1,276 @@
+import backtrader as bt
+import pandas as pd
+import yfinance as yf
+import json
+import os
+import shutil
+from datetime import datetime, timedelta
+from tradingagents.graph.trading_graph import TradingAgentsGraph
+from tradingagents.default_config import DEFAULT_CONFIG
+from dotenv import load_dotenv
+
+# Load environment variables
+load_dotenv()
+
+class TradingAgentsStrategy(bt.Strategy):
+    """Strategy that uses TradingAgents for decision making"""
+    
+    def __init__(self, trading_agent, ticker, backtest_config):
+        self.trading_agent = trading_agent
+        self.ticker = ticker
+        self.backtest_config = backtest_config
+        self.decisions = {}
+        self.trade_count = 0
+        self.data_checks = {}
+    
+    def next(self):
+        # Get current date
+        current_date = self.datas[0].datetime.date(0)
+        date_str = current_date.strftime("%Y-%m-%d")
+        
+        # Verify data range to avoid look-ahead bias
+        self.verify_data_range(date_str)
+        
+        # Get decision from TradingAgents
+        if date_str not in self.decisions:
+            print(f"Processing date: {date_str}")
+            try:
+                _, decision = self.trading_agent.propagate(self.ticker, date_str)
+                self.decisions[date_str] = decision
+                print(f"Decision: {decision}")
+            except Exception as e:
+                print(f"Error getting decision for {date_str}: {e}")
+                self.decisions[date_str] = "HOLD"
+        
+        decision = self.decisions[date_str]
+        
+        # Execute trade based on decision
+        if decision == "BUY" and not self.position:
+            # Buy with 100% of available cash
+            size = int(self.broker.getcash() / self.data.close[0])
+            if size > 0:
+                self.buy(size=size)
+                self.trade_count += 1
+                print(f"BUY {self.ticker} on {date_str} at ${self.data.close[0]:.2f}")
+        
+        elif decision == "SELL" and self.position:
+            # Sell all positions
+            self.sell(size=self.position.size)
+            self.trade_count += 1
+            print(f"SELL {self.ticker} on {date_str} at ${self.data.close[0]:.2f}")
+    
+    def verify_data_range(self, date_str):
+        """Verify that data range is correct to avoid look-ahead bias"""
+        current_date = datetime.strptime(date_str, "%Y-%m-%d")
+        
+        # Check if we already verified this date
+        if date_str in self.data_checks:
+            return
+        
+        # Verify data feed doesn't contain future data
+        data_end_date = self.datas[0].datetime.date(-1)
+        if data_end_date > current_date:
+            print(f"⚠️  Warning: Data feed contains future data beyond {date_str}")
+        
+        self.data_checks[date_str] = True
+
+def clean_cache():
+    """Clean cache to avoid look-ahead bias"""
+    print("\n=== Cleaning cache to avoid look-ahead bias ===")
+    
+    # Clean yfinance cache
+    yfinance_cache = "yfinance_cache"
+    if os.path.exists(yfinance_cache):
+        shutil.rmtree(yfinance_cache)
+        print(f"✓ Cleaned yfinance cache: {yfinance_cache}")
+    
+    # Clean dataflows cache
+    dataflows_cache = "dataflows/data_cache"
+    if os.path.exists(dataflows_cache):
+        shutil.rmtree(dataflows_cache)
+        print(f"✓ Cleaned dataflows cache: {dataflows_cache}")
+    
+    # Clean backtest results (optional)
+    # backtest_results = "backtest_results"
+    # if os.path.exists(backtest_results):
+    #     shutil.rmtree(backtest_results)
+    #     print(f"✓ Cleaned backtest results: {backtest_results}")
+
+def run_backtest(ticker, start_date, end_date, initial_cash=100000, clean_cache_flag=True):
+    """Run backtest for a given ticker and date range"""
+    
+    # Clean cache to avoid look-ahead bias
+    if clean_cache_flag:
+        clean_cache()
+    
+    # Create Cerebro engine
+    cerebro = bt.Cerebro()
+    
+    # Set initial cash
+    cerebro.broker.setcash(initial_cash)
+    
+    # Add strategy
+    config = DEFAULT_CONFIG.copy()
+    config["llm_provider"] = "openrouter"
+    config["deep_think_llm"] = "deepseek/deepseek-chat"
+    config["quick_think_llm"] = "openai/gpt-4o-mini"
+    config["max_debate_rounds"] = 2
+    
+    # Verify date range
+    start_dt = datetime.strptime(start_date, "%Y-%m-%d")
+    end_dt = datetime.strptime(end_date, "%Y-%m-%d")
+    
+    if start_dt >= end_dt:
+        raise ValueError("Start date must be before end date")
+    
+    if end_dt > datetime.now():
+        raise ValueError("End date cannot be in the future")
+    
+    trading_agent = TradingAgentsGraph(debug=False, config=config)
+    
+    backtest_config = {
+        "ticker": ticker,
+        "start_date": start_date,
+        "end_date": end_date,
+        "initial_cash": initial_cash,
+        "clean_cache": clean_cache_flag
+    }
+    
+    cerebro.addstrategy(TradingAgentsStrategy, trading_agent=trading_agent, ticker=ticker, backtest_config=backtest_config)
+    
+    # Get historical data from yfinance
+    print("\n=== Fetching historical data ===")
+    data = yf.download(ticker, start=start_date, end=end_date)
+    
+    # Verify data quality
+    if data.empty:
+        raise ValueError(f"No data found for {ticker} between {start_date} and {end_date}")
+    
+    print(f"✓ Data fetched: {len(data)} trading days")
+    print(f"✓ Date range: {data.index.min().date()} to {data.index.max().date()}")
+    
+    # Convert to backtrader data feed
+    data_feed = bt.feeds.PandasData(
+        dataname=data,
+        datetime=0,
+        high=1,
+        low=2,
+        open=3,
+        close=4,
+        volume=5,
+        openinterest=-1
+    )
+    
+    # Add data feed to cerebro
+    cerebro.adddata(data_feed, name=ticker)
+    
+    # Add analyzers
+    cerebro.addanalyzer(bt.analyzers.SharpeRatio, _name='sharpe')
+    cerebro.addanalyzer(bt.analyzers.DrawDown, _name='drawdown')
+    cerebro.addanalyzer(bt.analyzers.TradeAnalyzer, _name='trades')
+    cerebro.addanalyzer(bt.analyzers.AnnualReturn, _name='annual')
+    cerebro.addanalyzer(bt.analyzers.Returns, _name='returns')
+    cerebro.addanalyzer(bt.analyzers.PositionsValue, _name='positions')
+    
+    # Run backtest
+    print(f"\n=== Starting Backtest for {ticker} ===")
+    print(f"Date range: {start_date} to {end_date}")
+    print(f"Initial cash: ${initial_cash:.2f}")
+    print(f"LLM Provider: {config['llm_provider']}")
+    print(f"Models: Deep={config['deep_think_llm']}, Quick={config['quick_think_llm']}")
+    
+    results = cerebro.run()
+    
+    # Get results
+    strategy = results[0]
+    final_value = cerebro.broker.getvalue()
+    total_return = ((final_value - initial_cash) / initial_cash) * 100
+    
+    # Get analyzer results
+    sharpe = strategy.analyzers.sharpe.get_analysis()
+    drawdown = strategy.analyzers.drawdown.get_analysis()
+    trades = strategy.analyzers.trades.get_analysis()
+    annual = strategy.analyzers.annual.get_analysis()
+    returns = strategy.analyzers.returns.get_analysis()
+    
+    # Calculate additional metrics
+    total_trades = trades.get('total', {}).get('total', 0)
+    won_trades = trades.get('won', {}).get('total', 0)
+    win_rate = won_trades / max(total_trades, 1) * 100
+    
+    # Print results
+    print(f"\n=== Backtest Results ===")
+    print(f"Final portfolio value: ${final_value:.2f}")
+    print(f"Total return: {total_return:.2f}%")
+    print(f"Daily return: {returns.get('rnorm', 0) * 100:.4f}%")
+    print(f"Sharpe Ratio: {sharpe.get('sharperatio', 'N/A'):.2f}")
+    print(f"Max Drawdown: {drawdown.get('max', {}).get('drawdown', 'N/A'):.2f}%")
+    print(f"Total trades: {total_trades}")
+    print(f"Win rate: {win_rate:.2f}%")
+    print(f"Average trade duration: {trades.get('len', {}).get('average', 'N/A'):.1f} days")
+    
+    # Save results
+    save_results(ticker, start_date, end_date, {
+        "initial_cash": initial_cash,
+        "final_value": final_value,
+        "total_return": total_return,
+        "daily_return": returns.get('rnorm', 0),
+        "sharpe_ratio": sharpe.get('sharperatio', None),
+        "max_drawdown": drawdown.get('max', {}).get('drawdown', None),
+        "total_trades": total_trades,
+        "won_trades": won_trades,
+        "win_rate": win_rate,
+        "average_trade_duration": trades.get('len', {}).get('average', None),
+        "decisions": strategy.decisions,
+        "config": backtest_config
+    })
+    
+    # Plot results
+    print("\n=== Generating backtest chart ===")
+    cerebro.plot(style='candlestick')
+
+def save_results(ticker, start_date, end_date, results):
+    """Save backtest results to file"""
+    results_dir = f"backtest_results/{ticker}/"
+    os.makedirs(results_dir, exist_ok=True)
+    
+    filename = f"backtest_{start_date}_{end_date}.json"
+    filepath = os.path.join(results_dir, filename)
+    
+    with open(filepath, "w", encoding="utf-8") as f:
+        json.dump(results, f, indent=4, default=str)
+    
+    print(f"Results saved to: {filepath}")
+
+def run_multiple_backtests(ticker_list, start_date, end_date, initial_cash=100000):
+    """Run backtests for multiple tickers"""
+    all_results = {}
+    
+    for ticker in ticker_list:
+        print(f"\n{'='*60}")
+        print(f"Running backtest for {ticker}")
+        print(f"{'='*60}")
+        
+        try:
+            # Run backtest without cleaning cache for subsequent tickers
+            clean_cache_flag = (ticker == ticker_list[0])
+            run_backtest(ticker, start_date, end_date, initial_cash, clean_cache_flag)
+        except Exception as e:
+            print(f"Error running backtest for {ticker}: {e}")
+            all_results[ticker] = {"error": str(e)}
+    
+    return all_results
+
+if __name__ == "__main__":
+    # Define parameters
+    ticker = "NVDA"
+    start_date = "2024-01-01"
+    end_date = "2024-03-29"
+    initial_cash = 100000
+    
+    # Run backtest
+    run_backtest(ticker, start_date, end_date, initial_cash)
+    
+    # Example: Run multiple backtests
+    # tickers = ["NVDA", "AAPL", "MSFT"]
+    # run_multiple_backtests(tickers, start_date, end_date, initial_cash)

cli/main.py

@@ -6,8 +6,9 @@ from functools import wraps
 from rich.console import Console
 from dotenv import load_dotenv
 
-# Load environment variables from .env file
+# Load environment variables
 load_dotenv()
+load_dotenv(".env.enterprise", override=False)
 from rich.panel import Panel
 from rich.spinner import Spinner
 from rich.live import Live
@@ -79,7 +80,7 @@ class MessageBuffer:
         self.current_agent = None
         self.report_sections = {}
         self.selected_analysts = []
-        self._last_message_id = None
+        self._processed_message_ids = set()
 
     def init_for_analysis(self, selected_analysts):
         """Initialize agent status and report sections based on selected analysts.
@@ -114,7 +115,7 @@ class MessageBuffer:
         self.current_agent = None
         self.messages.clear()
         self.tool_calls.clear()
-        self._last_message_id = None
+        self._processed_message_ids.clear()
 
     def get_completed_reports_count(self):
         """Count reports that are finalized (their finalizing agent is completed).
@@ -1052,28 +1053,24 @@ def run_analysis():
         # Stream the analysis
         trace = []
         for chunk in graph.graph.stream(init_agent_state, **args):
-            # Process messages if present (skip duplicates via message ID)
-            if len(chunk["messages"]) > 0:
-                last_message = chunk["messages"][-1]
-                msg_id = getattr(last_message, "id", None)
+            # Process all messages in chunk, deduplicating by message ID
+            for message in chunk.get("messages", []):
+                msg_id = getattr(message, "id", None)
+                if msg_id is not None:
+                    if msg_id in message_buffer._processed_message_ids:
+                        continue
+                    message_buffer._processed_message_ids.add(msg_id)
 
-                if msg_id != message_buffer._last_message_id:
-                    message_buffer._last_message_id = msg_id
+                msg_type, content = classify_message_type(message)
+                if content and content.strip():
+                    message_buffer.add_message(msg_type, content)
 
-                    # Add message to buffer
-                    msg_type, content = classify_message_type(last_message)
-                    if content and content.strip():
-                        message_buffer.add_message(msg_type, content)
-
-                    # Handle tool calls
-                    if hasattr(last_message, "tool_calls") and last_message.tool_calls:
-                        for tool_call in last_message.tool_calls:
-                            if isinstance(tool_call, dict):
-                                message_buffer.add_tool_call(
-                                    tool_call["name"], tool_call["args"]
-                                )
-                            else:
-                                message_buffer.add_tool_call(tool_call.name, tool_call.args)
+                if hasattr(message, "tool_calls") and message.tool_calls:
+                    for tool_call in message.tool_calls:
+                        if isinstance(tool_call, dict):
+                            message_buffer.add_tool_call(tool_call["name"], tool_call["args"])
+                        else:
+                            message_buffer.add_tool_call(tool_call.name, tool_call.args)
 
             # Update analyst statuses based on report state (runs on every chunk)
             update_analyst_statuses(message_buffer, chunk)

cli/utils.py

@@ -174,17 +174,30 @@ def select_openrouter_model() -> str:
     return choice
 
 
-def select_shallow_thinking_agent(provider) -> str:
-    """Select shallow thinking llm engine using an interactive selection."""
+def _prompt_custom_model_id() -> str:
+    """Prompt user to type a custom model ID."""
+    return questionary.text(
+        "Enter model ID:",
+        validate=lambda x: len(x.strip()) > 0 or "Please enter a model ID.",
+    ).ask().strip()
 
+
+def _select_model(provider: str, mode: str) -> str:
+    """Select a model for the given provider and mode (quick/deep)."""
     if provider.lower() == "openrouter":
         return select_openrouter_model()
 
+    if provider.lower() == "azure":
+        return questionary.text(
+            f"Enter Azure deployment name ({mode}-thinking):",
+            validate=lambda x: len(x.strip()) > 0 or "Please enter a deployment name.",
+        ).ask().strip()
+
     choice = questionary.select(
-        "Select Your [Quick-Thinking LLM Engine]:",
+        f"Select Your [{mode.title()}-Thinking LLM Engine]:",
         choices=[
             questionary.Choice(display, value=value)
-            for display, value in get_model_options(provider, "quick")
+            for display, value in get_model_options(provider, mode)
         ],
         instruction="\n- Use arrow keys to navigate\n- Press Enter to select",
         style=questionary.Style(
@@ -197,58 +210,45 @@ def select_shallow_thinking_agent(provider) -> str:
     ).ask()
 
     if choice is None:
-        console.print(
-            "\n[red]No shallow thinking llm engine selected. Exiting...[/red]"
-        )
+        console.print(f"\n[red]No {mode} thinking llm engine selected. Exiting...[/red]")
         exit(1)
 
+    if choice == "custom":
+        return _prompt_custom_model_id()
+
     return choice
 
 
+def select_shallow_thinking_agent(provider) -> str:
+    """Select shallow thinking llm engine using an interactive selection."""
+    return _select_model(provider, "quick")
+
+
 def select_deep_thinking_agent(provider) -> str:
     """Select deep thinking llm engine using an interactive selection."""
-
-    if provider.lower() == "openrouter":
-        return select_openrouter_model()
-
-    choice = questionary.select(
-        "Select Your [Deep-Thinking LLM Engine]:",
-        choices=[
-            questionary.Choice(display, value=value)
-            for display, value in get_model_options(provider, "deep")
-        ],
-        instruction="\n- Use arrow keys to navigate\n- Press Enter to select",
-        style=questionary.Style(
-            [
-                ("selected", "fg:magenta noinherit"),
-                ("highlighted", "fg:magenta noinherit"),
-                ("pointer", "fg:magenta noinherit"),
-            ]
-        ),
-    ).ask()
-
-    if choice is None:
-        console.print("\n[red]No deep thinking llm engine selected. Exiting...[/red]")
-        exit(1)
-
-    return choice
+    return _select_model(provider, "deep")
 
 def select_llm_provider() -> tuple[str, str | None]:
     """Select the LLM provider and its API endpoint."""
-    BASE_URLS = [
-        ("OpenAI", "https://api.openai.com/v1"),
-        ("Google", None),  # google-genai SDK manages its own endpoint
-        ("Anthropic", "https://api.anthropic.com/"),
-        ("xAI", "https://api.x.ai/v1"),
-        ("Openrouter", "https://openrouter.ai/api/v1"),
-        ("Ollama", "http://localhost:11434/v1"),
+    # (display_name, provider_key, base_url)
+    PROVIDERS = [
+        ("OpenAI", "openai", "https://api.openai.com/v1"),
+        ("Google", "google", None),
+        ("Anthropic", "anthropic", "https://api.anthropic.com/"),
+        ("xAI", "xai", "https://api.x.ai/v1"),
+        ("DeepSeek", "deepseek", "https://api.deepseek.com"),
+        ("Qwen", "qwen", "https://dashscope.aliyuncs.com/compatible-mode/v1"),
+        ("GLM", "glm", "https://open.bigmodel.cn/api/paas/v4/"),
+        ("OpenRouter", "openrouter", "https://openrouter.ai/api/v1"),
+        ("Azure OpenAI", "azure", None),
+        ("Ollama", "ollama", "http://localhost:11434/v1"),
     ]
-    
+
     choice = questionary.select(
         "Select your LLM Provider:",
         choices=[
-            questionary.Choice(display, value=(display, value))
-            for display, value in BASE_URLS
+            questionary.Choice(display, value=(provider_key, url))
+            for display, provider_key, url in PROVIDERS
         ],
         instruction="\n- Use arrow keys to navigate\n- Press Enter to select",
         style=questionary.Style(
@@ -261,13 +261,11 @@ def select_llm_provider() -> tuple[str, str | None]:
     ).ask()
     
     if choice is None:
-        console.print("\n[red]no OpenAI backend selected. Exiting...[/red]")
+        console.print("\n[red]No LLM provider selected. Exiting...[/red]")
         exit(1)
-    
-    display_name, url = choice
-    print(f"You selected: {display_name}\tURL: {url}")
 
-    return display_name, url
+    provider, url = choice
+    return provider, url
 
 
 def ask_openai_reasoning_effort() -> str:

docker-compose.yml

@@ -4,7 +4,7 @@ services:
     env_file:
       - .env
     volumes:
-      - ./results:/home/appuser/app/results
+      - tradingagents_data:/home/appuser/.tradingagents
     tty: true
     stdin_open: true
 
@@ -22,7 +22,7 @@ services:
     environment:
       - LLM_PROVIDER=ollama
     volumes:
-      - ./results:/home/appuser/app/results
+      - tradingagents_data:/home/appuser/.tradingagents
     depends_on:
       - ollama
     tty: true
@@ -31,4 +31,5 @@ services:
       - ollama
 
 volumes:
+  tradingagents_data:
   ollama_data:

tradingagents/agents/utils/memory.py

@@ -78,7 +78,7 @@ class FinancialSituationMemory:
 
         # Build results
         results = []
-        max_score = max(scores) if max(scores) > 0 else 1  # Normalize scores
+        max_score = float(scores.max()) if len(scores) > 0 and scores.max() > 0 else 1.0
 
         for idx in top_indices:
             # Normalize score to 0-1 range for consistency

tradingagents/default_config.py

@@ -1,12 +1,11 @@
 import os
 
+_TRADINGAGENTS_HOME = os.path.join(os.path.expanduser("~"), ".tradingagents")
+
 DEFAULT_CONFIG = {
     "project_dir": os.path.abspath(os.path.join(os.path.dirname(__file__), ".")),
-    "results_dir": os.getenv("TRADINGAGENTS_RESULTS_DIR", "./results"),
-    "data_cache_dir": os.path.join(
-        os.path.abspath(os.path.join(os.path.dirname(__file__), ".")),
-        "dataflows/data_cache",
-    ),
+    "results_dir": os.getenv("TRADINGAGENTS_RESULTS_DIR", os.path.join(_TRADINGAGENTS_HOME, "logs")),
+    "data_cache_dir": os.getenv("TRADINGAGENTS_CACHE_DIR", os.path.join(_TRADINGAGENTS_HOME, "cache")),
     # LLM settings
     "llm_provider": "openai",
     "deep_think_llm": "gpt-5.4",

tradingagents/graph/trading_graph.py

@@ -66,10 +66,8 @@ class TradingAgentsGraph:
         set_config(self.config)
 
         # Create necessary directories
-        os.makedirs(
-            os.path.join(self.config["project_dir"], "dataflows/data_cache"),
-            exist_ok=True,
-        )
+        os.makedirs(self.config["data_cache_dir"], exist_ok=True)
+        os.makedirs(self.config["results_dir"], exist_ok=True)
 
         # Initialize LLMs with provider-specific thinking configuration
         llm_kwargs = self._get_provider_kwargs()

tradingagents/llm_clients/azure_client.py

@@ -0,0 +1,52 @@
+import os
+from typing import Any, Optional
+
+from langchain_openai import AzureChatOpenAI
+
+from .base_client import BaseLLMClient, normalize_content
+from .validators import validate_model
+
+_PASSTHROUGH_KWARGS = (
+    "timeout", "max_retries", "api_key", "reasoning_effort",
+    "callbacks", "http_client", "http_async_client",
+)
+
+
+class NormalizedAzureChatOpenAI(AzureChatOpenAI):
+    """AzureChatOpenAI with normalized content output."""
+
+    def invoke(self, input, config=None, **kwargs):
+        return normalize_content(super().invoke(input, config, **kwargs))
+
+
+class AzureOpenAIClient(BaseLLMClient):
+    """Client for Azure OpenAI deployments.
+
+    Requires environment variables:
+        AZURE_OPENAI_API_KEY: API key
+        AZURE_OPENAI_ENDPOINT: Endpoint URL (e.g. https://<resource>.openai.azure.com/)
+        AZURE_OPENAI_DEPLOYMENT_NAME: Deployment name
+        OPENAI_API_VERSION: API version (e.g. 2025-03-01-preview)
+    """
+
+    def __init__(self, model: str, base_url: Optional[str] = None, **kwargs):
+        super().__init__(model, base_url, **kwargs)
+
+    def get_llm(self) -> Any:
+        """Return configured AzureChatOpenAI instance."""
+        self.warn_if_unknown_model()
+
+        llm_kwargs = {
+            "model": self.model,
+            "azure_deployment": os.environ.get("AZURE_OPENAI_DEPLOYMENT_NAME", self.model),
+        }
+
+        for key in _PASSTHROUGH_KWARGS:
+            if key in self.kwargs:
+                llm_kwargs[key] = self.kwargs[key]
+
+        return NormalizedAzureChatOpenAI(**llm_kwargs)
+
+    def validate_model(self) -> bool:
+        """Azure accepts any deployed model name."""
+        return True

tradingagents/llm_clients/factory.py

@@ -4,6 +4,12 @@ from .base_client import BaseLLMClient
 from .openai_client import OpenAIClient
 from .anthropic_client import AnthropicClient
 from .google_client import GoogleClient
+from .azure_client import AzureOpenAIClient
+
+# Providers that use the OpenAI-compatible chat completions API
+_OPENAI_COMPATIBLE = (
+    "openai", "xai", "deepseek", "qwen", "glm", "ollama", "openrouter",
+)
 
 
 def create_llm_client(
@@ -15,16 +21,10 @@ def create_llm_client(
     """Create an LLM client for the specified provider.
 
     Args:
-        provider: LLM provider (openai, anthropic, google, xai, ollama, openrouter)
+        provider: LLM provider name
         model: Model name/identifier
         base_url: Optional base URL for API endpoint
         **kwargs: Additional provider-specific arguments
-            - http_client: Custom httpx.Client for SSL proxy or certificate customization
-            - http_async_client: Custom httpx.AsyncClient for async operations
-            - timeout: Request timeout in seconds
-            - max_retries: Maximum retry attempts
-            - api_key: API key for the provider
-            - callbacks: LangChain callbacks
 
     Returns:
         Configured BaseLLMClient instance
@@ -34,16 +34,16 @@ def create_llm_client(
     """
     provider_lower = provider.lower()
 
-    if provider_lower in ("openai", "ollama", "openrouter"):
+    if provider_lower in _OPENAI_COMPATIBLE:
         return OpenAIClient(model, base_url, provider=provider_lower, **kwargs)
 
-    if provider_lower == "xai":
-        return OpenAIClient(model, base_url, provider="xai", **kwargs)
-
     if provider_lower == "anthropic":
         return AnthropicClient(model, base_url, **kwargs)
 
     if provider_lower == "google":
         return GoogleClient(model, base_url, **kwargs)
 
+    if provider_lower == "azure":
+        return AzureOpenAIClient(model, base_url, **kwargs)
+
     raise ValueError(f"Unsupported LLM provider: {provider}")

tradingagents/llm_clients/model_catalog.py

@@ -63,8 +63,43 @@ MODEL_OPTIONS: ProviderModeOptions = {
             ("Grok 4.1 Fast (Non-Reasoning) - Speed optimized, 2M ctx", "grok-4-1-fast-non-reasoning"),
         ],
     },
-    # OpenRouter models are fetched dynamically at CLI runtime.
-    # No static entries needed; any model ID is accepted by the validator.
+    "deepseek": {
+        "quick": [
+            ("DeepSeek V3.2", "deepseek-chat"),
+            ("Custom model ID", "custom"),
+        ],
+        "deep": [
+            ("DeepSeek V3.2 (thinking)", "deepseek-reasoner"),
+            ("DeepSeek V3.2", "deepseek-chat"),
+            ("Custom model ID", "custom"),
+        ],
+    },
+    "qwen": {
+        "quick": [
+            ("Qwen 3.5 Flash", "qwen3.5-flash"),
+            ("Qwen Plus", "qwen-plus"),
+            ("Custom model ID", "custom"),
+        ],
+        "deep": [
+            ("Qwen 3.6 Plus", "qwen3.6-plus"),
+            ("Qwen 3.5 Plus", "qwen3.5-plus"),
+            ("Qwen 3 Max", "qwen3-max"),
+            ("Custom model ID", "custom"),
+        ],
+    },
+    "glm": {
+        "quick": [
+            ("GLM-4.7", "glm-4.7"),
+            ("GLM-5", "glm-5"),
+            ("Custom model ID", "custom"),
+        ],
+        "deep": [
+            ("GLM-5.1", "glm-5.1"),
+            ("GLM-5", "glm-5"),
+            ("Custom model ID", "custom"),
+        ],
+    },
+    # OpenRouter: fetched dynamically. Azure: any deployed model name.
     "ollama": {
         "quick": [
             ("Qwen3:latest (8B, local)", "qwen3:latest"),

tradingagents/llm_clients/openai_client.py

@@ -27,6 +27,9 @@ _PASSTHROUGH_KWARGS = (
 # Provider base URLs and API key env vars
 _PROVIDER_CONFIG = {
     "xai": ("https://api.x.ai/v1", "XAI_API_KEY"),
+    "deepseek": ("https://api.deepseek.com", "DEEPSEEK_API_KEY"),
+    "qwen": ("https://dashscope-intl.aliyuncs.com/compatible-mode/v1", "DASHSCOPE_API_KEY"),
+    "glm": ("https://api.z.ai/api/paas/v4/", "ZHIPU_API_KEY"),
     "openrouter": ("https://openrouter.ai/api/v1", "OPENROUTER_API_KEY"),
     "ollama": ("http://localhost:11434/v1", None),
 }