diff --git a/.env.example b/.env.example
index 1328b838..e547b91c 100644
--- a/.env.example
+++ b/.env.example
@@ -1,6 +1,22 @@
-# LLM Providers (set the one you use)
+# ─── Cloud Providers ──────────────────────────────────────────────────────────
 OPENAI_API_KEY=
 GOOGLE_API_KEY=
 ANTHROPIC_API_KEY=
 XAI_API_KEY=
 OPENROUTER_API_KEY=
+
+# ─── Local LLM via llama.cpp ──────────────────────────────────────────────────
+# 1. Start llama-server:
+#    llama-server --model ~/models/my-model.gguf --port 8081 --host 0.0.0.0
+# 2. Copy this file to .env and uncomment + fill in the values below
+# 3. Find your model ID with: curl http://localhost:8081/v1/models
+
+# LLM_PROVIDER=llamacpp
+# BACKEND_URL=http://localhost:8081/v1
+# DEEP_THINK_LLM=your-model-name-here
+# QUICK_THINK_LLM=your-model-name-here
+
+# ─── Any OpenAI-compatible local server (LM Studio, vLLM, etc.) ───────────────
+# LLM_PROVIDER=openai
+# BACKEND_URL=http://localhost:1234/v1
+# OPENAI_API_KEY=dummy
diff --git a/.gitignore b/.gitignore
index 9a2904a9..1bb7fd64 100644
--- a/.gitignore
+++ b/.gitignore
@@ -217,3 +217,4 @@ __marimo__/
 
 # Cache
 **/data_cache/
+!.env.example
diff --git a/tradingagents/default_config.py b/tradingagents/default_config.py
index ecf0dc29..f553ff8e 100644
--- a/tradingagents/default_config.py
+++ b/tradingagents/default_config.py
@@ -1,5 +1,11 @@
 import os
 
+try:
+    from dotenv import load_dotenv
+    load_dotenv()
+except ImportError:
+    pass  # python-dotenv optional — falls back to environment variables
+
 DEFAULT_CONFIG = {
     "project_dir": os.path.abspath(os.path.join(os.path.dirname(__file__), ".")),
     "results_dir": os.getenv("TRADINGAGENTS_RESULTS_DIR", "./results"),
@@ -8,10 +14,10 @@ DEFAULT_CONFIG = {
         "dataflows/data_cache",
     ),
     # LLM settings
-    "llm_provider": "openai",
-    "deep_think_llm": "gpt-5.2",
-    "quick_think_llm": "gpt-5-mini",
-    "backend_url": "https://api.openai.com/v1",
+    "llm_provider": os.environ.get("LLM_PROVIDER", "openai"),
+    "deep_think_llm": os.environ.get("DEEP_THINK_LLM", "gpt-5.2"),
+    "quick_think_llm": os.environ.get("QUICK_THINK_LLM", "gpt-5-mini"),
+    "backend_url": os.environ.get("BACKEND_URL", os.environ.get("LLAMACPP_BASE_URL", "https://api.openai.com/v1")),
     # Provider-specific thinking configuration
     "google_thinking_level": None,      # "high", "minimal", etc.
     "openai_reasoning_effort": None,    # "medium", "high", "low"
diff --git a/tradingagents/llm_clients/factory.py b/tradingagents/llm_clients/factory.py
index 93c2a7d3..af593d9f 100644
--- a/tradingagents/llm_clients/factory.py
+++ b/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", "llamacpp"):
         return OpenAIClient(model, base_url, provider=provider_lower, **kwargs)
 
     if provider_lower == "xai":
diff --git a/tradingagents/llm_clients/openai_client.py b/tradingagents/llm_clients/openai_client.py
index 4605c1f9..da16f3a9 100644
--- a/tradingagents/llm_clients/openai_client.py
+++ b/tradingagents/llm_clients/openai_client.py
@@ -56,6 +56,10 @@ class OpenAIClient(BaseLLMClient):
         elif self.provider == "ollama":
             llm_kwargs["base_url"] = "http://localhost:11434/v1"
             llm_kwargs["api_key"] = "ollama"  # Ollama doesn't require auth
+        elif self.provider == "llamacpp":
+            base_url = self.base_url or os.environ.get("LLAMACPP_BASE_URL", "http://localhost:8080/v1")
+            llm_kwargs["base_url"] = base_url
+            llm_kwargs["api_key"] = "no-key-needed"  # llama-server doesn't require auth
         elif self.base_url:
             llm_kwargs["base_url"] = self.base_url
 
diff --git a/tradingagents/llm_clients/validators.py b/tradingagents/llm_clients/validators.py
index 1e2388b3..0c201cec 100644
--- a/tradingagents/llm_clients/validators.py
+++ b/tradingagents/llm_clients/validators.py
@@ -58,7 +58,7 @@ def validate_model(provider: str, model: str) -> bool:
     """
     provider_lower = provider.lower()
 
-    if provider_lower in ("ollama", "openrouter"):
+    if provider_lower in ("ollama", "openrouter", "llamacpp"):
         return True
 
     if provider_lower not in VALID_MODELS: