A reactive notebook for Python — run reproducible experiments, query with SQL, execute as a script, deploy as an app, and version with git. Stored as pure Python. All in a modern, AI-native editor.
Python artificial-intelligencedagdata-sciencedata-visualizationdataflowdeveloper-toolsmachine-learningnotebookspipelinepythonreactivesqlweb-app
# /// script
# requires-python = ">=3.11"
# dependencies = [
# "marimo",
# ]
# ///
import marimo
__generated_with = "0.20.2"
app = marimo.App(width="medium")
@app.cell
def _():
import marimo as mo
return (mo,)
@app.cell(hide_code=True)
def _(mo):
mo.md("""
# Custom chatbot
This example shows how to make a custom chatbot: just supply a function that takes two arguments,
`messages` and `config`, and returns the chatbot's response. This response can be any object; it
doesn't have to be a string!
""")
return
@app.cell
def _(mo):
def simple_echo_model(messages, config):
"""This chatbot echoes what the user says."""
# messages is a list of chatbot messages
message = messages[-1]
# Each message has two fields:
# message.role, which may be "user", "assistant", or "system"
# message.content: the content of the message
return f"You said: {messages[-1].content}!"
chatbot = mo.ui.chat(
simple_echo_model,
prompts=["Hello", "How are you?"],
show_configuration_controls=False,
)
chatbot
return (chatbot,)
@app.cell
def _(mo):
mo.md("""
Access the chatbot's historical messages with `chatbot.value`.
""")
return
@app.cell
def _(chatbot):
# chatbot.value is the list of chat messages
chatbot.value
return
@app.cell
def _(mo):
def marimo_chatbot(messages, config):
"""Returns marimo elements"""
message = messages[-1]
return mo.vstack(
[
mo.md("## Response"),
mo.ui.table([1, 2, 3]),
mo.ui.slider(1, 10, label="slider"),
]
)
mo.ui.chat(marimo_chatbot)
return
if __name__ == "__main__":
app.run()