Setup:

Install langchain-snowflake and configure Snowflake connection.

.. code-block:: bash

pip install -U langchain-snowflake

Key init args — completion params: model: str Name of Snowflake Cortex model to use (e.g., 'llama3.1-70b', 'mistral-large2') temperature: float Sampling temperature (0.0 to 1.0) max_tokens: Optional[int] Max number of tokens to generate (default: 4096)

Key init args — client params: session: Optional[Session] Active Snowflake session. If not provided, will create from connection params. account: Optional[str] Snowflake account identifier user: Optional[str] Snowflake username password: Optional[SecretStr] Snowflake password warehouse: Optional[str] Snowflake warehouse to use database: Optional[str] Snowflake database to use schema: Optional[str] Snowflake schema to use request_timeout: int Request timeout in seconds for API calls (default: 300) verify_ssl: bool Whether to verify SSL certificates (default: True)

Instantiate:

.. code-block:: python

from .. import ChatSnowflake

Using existing session

llm = ChatSnowflake( model="llama3.1-70b", session=session, temperature=0.1, max_tokens=1000 )

Using connection parameters with network configuration

llm = ChatSnowflake( model="mistral-large2", account="your-account", user="your-user", password="your-password", warehouse="your-warehouse", temperature=0.0, request_timeout=600, # 10 minutes for long-running operations verify_ssl=True # Always verify SSL in production )

Invoke:

.. code-block:: python

messages = [ ("system", "You are a helpful assistant."), ("human", "What is the capital of France?"), ] response = llm.invoke(messages) print(response.content)

Stream:

.. code-block:: python

for chunk in llm.stream(messages): print(chunk.content, end="", flush=True)

Async:

.. code-block:: python

response = await llm.ainvoke(messages) async for chunk in llm.astream(messages): print(chunk.content, end="", flush=True)

Tool calling:

.. code-block:: python

@tool def get_weather(city: str) -> str: '''Get weather for a city.''' return f"The weather in {city} is 72°F and sunny."

llm_with_tools = llm.bind_tools([get_weather]) messages = [("human", "What's the weather in Paris?")] response = llm_with_tools.invoke(messages)

Structured output:

.. code-block:: python

from typing import Literal from pydantic import BaseModel

class Sentiment(BaseModel): sentiment: Literal["positive", "negative", "neutral"] confidence: float

structured_llm = llm.with_structured_output(Sentiment) result = structured_llm.invoke("I love this product!") print(result.sentiment, result.confidence)

Response metadata:

.. code-block:: python

response = llm.invoke(messages) print(response.response_metadata)

{'model': 'llama3.1-70b', 'usage': {'prompt_tokens': 10, 'completion_tokens': 5}}