ai-agents-llm-provider-configu-need-to-use-different-llms-for-different-reasoning-07f6c2e3

THEN Configure separate models for complex reasoning (deep_think_llm) and quick tasks (quick_think_llm) via a config dictionary. Supports multiple providers: OpenAI, Google, Anthropic, xAI, DeepSeek, Qwen, GLM, OpenRouter, Ollama, Azure. Set the provider with 'llm_provider' key. For local models, use 'ollama' and run with Docker Compose ollama profile.

config = DEFAULT_CONFIG.copy()
config["llm_provider"] = "openai"
config["deep_think_llm"] = "gpt-5.4"
config["quick_think_llm"] = "gpt-5.4-mini"
ta = TradingAgentsGraph(debug=True, config=config)
_, decision = ta.propagate("AAPL", "2026-02-01")