Skip to main content


Baseten provides all the infrastructure you need to deploy and serve ML models performantly, scalably, and cost-efficiently.

This example demonstrates using Langchain with models deployed on Baseten.


To run this notebook, you'll need a Baseten account and an API key.

You'll also need to install the Baseten Python package:

pip install baseten
import baseten


Single model call

First, you'll need to deploy a model to Baseten.

You can deploy foundation models like WizardLM and Alpaca with one click from the Baseten model library or if you have your own model, deploy it with this tutorial.

In this example, we'll work with WizardLM. Deploy WizardLM here and follow along with the deployed model's version ID.

from langchain.llms import Baseten

API Reference:

# Load the model
wizardlm = Baseten(model="MODEL_VERSION_ID", verbose=True)
# Prompt the model

wizardlm("What is the difference between a Wizard and a Sorcerer?")

Chained model calls

We can chain together multiple calls to one or multiple models, which is the whole point of Langchain!

This example uses WizardLM to plan a meal with an entree, three sides, and an alcoholic and non-alcoholic beverage pairing.

from langchain.chains import SimpleSequentialChain
from langchain.prompts import PromptTemplate
from langchain.chains import LLMChain
# Build the first link in the chain

prompt = PromptTemplate(
template="Name a complex entree for a {cuisine} dinner. Respond with just the name of a single dish.",

link_one = LLMChain(llm=wizardlm, prompt=prompt)
# Build the second link in the chain

prompt = PromptTemplate(
template="What are three sides that would go with {entree}. Respond with only a list of the sides.",

link_two = LLMChain(llm=wizardlm, prompt=prompt)
# Build the third link in the chain

prompt = PromptTemplate(
template="What is one alcoholic and one non-alcoholic beverage that would go well with this list of sides: {sides}. Respond with only the names of the beverages.",

link_three = LLMChain(llm=wizardlm, prompt=prompt)
# Run the full chain!

menu_maker = SimpleSequentialChain(
chains=[link_one, link_two, link_three], verbose=True
)"South Indian")