Add Cerebras as LLM provider using OpenAI-compatible API

- Register Cerebras base URL and API key env var in openai_client.py
- Route cerebras provider through OpenAIClient in factory.py
- Add available Cerebras models to validators.py
- Add Cerebras to CLI provider menu and model selection in cli/utils.py
- Add cerebras override in cli/main.py config
- Set main.py to use Cerebras llama3.1-8b (llama3.1-8b has working tool calling)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

cli/main.py

@@ -927,6 +927,10 @@ def run_analysis():
     config["deep_think_llm"] = selections["deep_thinker"]
     config["backend_url"] = selections["backend_url"]
     config["llm_provider"] = selections["llm_provider"].lower()
+    # Override with Cerebras
+    config["llm_provider"] = "cerebras"
+    config["deep_think_llm"] = "llama-4-scout-17b-16e-instruct"
+    config["quick_think_llm"] = "llama-4-scout-17b-16e-instruct"
     # Provider-specific thinking configuration
     config["google_thinking_level"] = selections.get("google_thinking_level")
     config["openai_reasoning_effort"] = selections.get("openai_reasoning_effort")

cli/utils.py

@@ -171,6 +171,10 @@ def select_shallow_thinking_agent(provider) -> str:
             ("GPT-OSS:latest (20B, local)", "gpt-oss:latest"),
             ("GLM-4.7-Flash:latest (30B, local)", "glm-4.7-flash:latest"),
         ],
+        "cerebras": [
+            ("Llama 3.1 8B - Fastest, lightweight", "llama3.1-8b"),
+            ("Qwen 3 235B - Most capable", "qwen-3-235b-a22b-instruct-2507"),
+        ],
     }
 
     choice = questionary.select(
@@ -238,6 +242,10 @@ def select_deep_thinking_agent(provider) -> str:
             ("GPT-OSS:latest (20B, local)", "gpt-oss:latest"),
             ("Qwen3:latest (8B, local)", "qwen3:latest"),
         ],
+        "cerebras": [
+            ("Qwen 3 235B - Most capable", "qwen-3-235b-a22b-instruct-2507"),
+            ("Llama 3.1 8B - Fastest, lightweight", "llama3.1-8b"),
+        ],
     }
 
     choice = questionary.select(
@@ -272,6 +280,7 @@ def select_llm_provider() -> tuple[str, str]:
         ("xAI", "https://api.x.ai/v1"),
         ("Openrouter", "https://openrouter.ai/api/v1"),
         ("Ollama", "http://localhost:11434/v1"),
+        ("Cerebras", "https://api.cerebras.ai/v1"),
     ]
     
     choice = questionary.select(

main.py

@@ -8,9 +8,10 @@ load_dotenv()
 
 # Create a custom config
 config = DEFAULT_CONFIG.copy()
-config["deep_think_llm"] = "gpt-5-mini"  # Use a different model
-config["quick_think_llm"] = "gpt-5-mini"  # Use a different model
-config["max_debate_rounds"] = 1  # Increase debate rounds
+config["llm_provider"] = "cerebras"
+config["deep_think_llm"] = "llama3.1-8b"
+config["quick_think_llm"] = "llama3.1-8b"
+config["max_debate_rounds"] = 1
 
 # Configure data vendors (default uses yfinance, no extra API keys needed)
 config["data_vendors"] = {

tradingagents/llm_clients/factory.py

@@ -34,7 +34,7 @@ def create_llm_client(
     """
     provider_lower = provider.lower()
 
-    if provider_lower in ("openai", "ollama", "openrouter"):
+    if provider_lower in ("openai", "ollama", "openrouter", "cerebras"):
         return OpenAIClient(model, base_url, provider=provider_lower, **kwargs)
 
     if provider_lower == "xai":

tradingagents/llm_clients/openai_client.py

@@ -29,6 +29,7 @@ _PROVIDER_CONFIG = {
     "xai": ("https://api.x.ai/v1", "XAI_API_KEY"),
     "openrouter": ("https://openrouter.ai/api/v1", "OPENROUTER_API_KEY"),
     "ollama": ("http://localhost:11434/v1", None),
+    "cerebras": ("https://api.cerebras.ai/v1", "CEREBRAS_API_KEY"),
 }
 
 

tradingagents/llm_clients/validators.py

@@ -48,6 +48,10 @@ VALID_MODELS = {
         "grok-4-fast-reasoning",
         "grok-4-fast-non-reasoning",
     ],
+    "cerebras": [
+        "qwen-3-235b-a22b-instruct-2507",
+        "llama3.1-8b",
+    ],
 }