Models

• The models (LLMs from an AI provider) can be utilized in 2 ways:

• With agents

  • Using langchain.agents.create_agent

• Standalone: called outside of the agent loop for tasks like text generation, classification, etc

  • Using langchain.chat_models.init_chat_model
  • Using provider-specific classes langchain_google_genai.ChatGoogleGenerativeAI, langchain_openai.ChatOpenAI, etc

pip install langchain[openai]
pip install langchain[google-genai]
pip install langchain[ollama]

uv add langchain-openai
uv add langchain-google-genai
uv add langchain-ollama