feat: allow custom model and API URL input for LLM selection in CLI utilities

cli/utils.py

@@ -125,6 +125,25 @@ def select_research_depth() -> int:
 def select_shallow_thinking_agent(provider) -> str:
     """Select shallow thinking llm engine using an interactive selection."""
 
+    # if provider is custom, let user input model name
+    if provider.lower() == "custom":
+        model = questionary.text(
+            "Enter your custom Quick-Thinking model name:",
+            validate=lambda x: len(
+                x.strip()) > 0 or "Please enter a valid model name.",
+            style=questionary.Style(
+                [
+                    ("text", "fg:green"),
+                    ("highlighted", "noinherit"),
+                ]
+            ),
+        ).ask()
+
+        if not model:
+            console.print("\n[red]No model name provided. Exiting...[/red]")
+            exit(1)
+
+        return model.strip()
     # Define shallow thinking llm engine options with their corresponding model names
     SHALLOW_AGENT_OPTIONS = {
         "openai": [
@@ -183,6 +202,26 @@ def select_shallow_thinking_agent(provider) -> str:
 def select_deep_thinking_agent(provider) -> str:
     """Select deep thinking llm engine using an interactive selection."""
 
+    # if provider is custom, let user input model name
+    if provider.lower() == "custom":
+        model = questionary.text(
+            "Enter your custom Deep-Thinking model name:",
+            validate=lambda x: len(
+                x.strip()) > 0 or "Please enter a valid model name.",
+            style=questionary.Style(
+                [
+                    ("text", "fg:green"),
+                    ("highlighted", "noinherit"),
+                ]
+            ),
+        ).ask()
+
+        if not model:
+            console.print("\n[red]No model name provided. Exiting...[/red]")
+            exit(1)
+
+        return model.strip()
+
     # Define deep thinking llm engine options with their corresponding model names
     DEEP_AGENT_OPTIONS = {
         "openai": [
@@ -247,7 +286,8 @@ def select_llm_provider() -> tuple[str, str]:
         ("Anthropic", "https://api.anthropic.com/"),
         ("Google", "https://generativelanguage.googleapis.com/v1"),
         ("Openrouter", "https://openrouter.ai/api/v1"),
-        ("Ollama", "http://localhost:11434/v1"),        
+        ("Ollama", "http://localhost:11434/v1"),
+        ("Custom (Enter your custom backend URL)", "custom"),        
     ]
     
     choice = questionary.select(
@@ -269,7 +309,31 @@ def select_llm_provider() -> tuple[str, str]:
     if choice is None:
         console.print("\n[red]no OpenAI backend selected. Exiting...[/red]")
         exit(1)
-    
+
+    # if custom, let user input custom url and api key
+    if choice[1] == "custom":
+        custom_url = questionary.text(
+            "Enter your custom API base URL:",
+            validate=lambda x: len(x.strip()) > 0 or "Please enter a valid backend URL.").ask()
+
+        if not custom_url:
+            console.print("\n[red]No URL provided. Exiting...[/red]")
+            exit(1)
+
+        api_key = questionary.text(
+            "Enter your API key:",
+            validate=lambda x: len(x.strip()) > 0 or "Please enter a valid API key.").ask()
+
+        if not api_key:
+            console.print("\n[red]No API key provided. Exiting...[/red]")
+            exit(1)
+
+        # set api key to environment variable
+        import os
+        os.environ["OPENAI_API_KEY"] = api_key.strip()
+
+        return "Custom", custom_url.strip()
+
     display_name, url = choice
     print(f"You selected: {display_name}\tURL: {url}")
     

tradingagents/graph/trading_graph.py

@@ -58,7 +58,7 @@ class TradingAgentsGraph:
         )
 
         # Initialize LLMs
-        if self.config["llm_provider"].lower() == "openai" or self.config["llm_provider"] == "ollama" or self.config["llm_provider"] == "openrouter":
+        if self.config["llm_provider"].lower() in ["openai", "ollama", "openrouter", "custom"]:
             self.deep_thinking_llm = ChatOpenAI(model=self.config["deep_think_llm"], base_url=self.config["backend_url"])
             self.quick_thinking_llm = ChatOpenAI(model=self.config["quick_think_llm"], base_url=self.config["backend_url"])
         elif self.config["llm_provider"].lower() == "anthropic":