langchain/libs/partners/ai21

langchain-ai21

This package contains the LangChain integrations for AI21 through their AI21 SDK.

Installation and Setup

• Install the AI21 partner package

pip install langchain-ai21

• Get an AI21 api key and set it as an environment variable (AI21_API_KEY)

Chat Models

This package contains the ChatAI21 class, which is the recommended way to interface with AI21 Chat models.

To use, install the requirements, and configure your environment.

export AI21_API_KEY=your-api-key

Then initialize

from langchain_core.messages import HumanMessage
from langchain_ai21.chat_models import ChatAI21

chat = ChatAI21(model="j2-ultra")
messages = [HumanMessage(content="Hello from AI21")]
chat.invoke(messages)

LLMs

You can use AI21's generative AI models as Langchain LLMs:

from langchain.prompts import PromptTemplate
from langchain_ai21 import AI21LLM

llm = AI21LLM(model="j2-ultra")

template = """Question: {question}

Answer: Let's think step by step."""
prompt = PromptTemplate.from_template(template)

chain = prompt | llm

question = "Which scientist discovered relativity?"
print(chain.invoke({"question": question}))

Embeddings

You can use AI21's embeddings models as:

Query

from langchain_ai21 import AI21Embeddings

embeddings = AI21Embeddings()
embeddings.embed_query("Hello! This is some query")

Document

from langchain_ai21 import AI21Embeddings

embeddings = AI21Embeddings()
embeddings.embed_documents(["Hello! This is document 1", "And this is document 2!"])

Task Specific Models

Contextual Answers

You can use AI21's contextual answers model to receives text or document, serving as a context, and a question and returns an answer based entirely on this context.

This means that if the answer to your question is not in the document, the model will indicate it (instead of providing a false answer)

from langchain_ai21 import AI21ContextualAnswers

tsm = AI21ContextualAnswers()

response = tsm.invoke(input={"context": "Your context", "question": "Your question"})

You can also use it with chains and output parsers and vector DBs:

from langchain_ai21 import AI21ContextualAnswers
from langchain_core.output_parsers import StrOutputParser

tsm = AI21ContextualAnswers()
chain = tsm | StrOutputParser()

response = chain.invoke(
    {"context": "Your context", "question": "Your question"},
)