Course navigation
Chat History & MemoryLesson 7 of 8

Introduction to Streamlit

After MongoDB Chat Message History, this lesson moves from terminal scripts to a browser UI. Streamlit turns Python into a local web page — install it, run a starter app, and note how the script reruns on each click. LangChain wiring is in the next lesson.

Before you start

Finish the history lessons through MongoDB first. This page has no LangChain imports — only Streamlit layout and st.chat_input.

How it works

You call widgets in plain Python — st.title(), st.chat_input() — and Streamlit serves the page at localhost:8501. Each interaction reruns the script from top to bottom.

On each click:

1. Browser sends the click to Streamlit server
2. Streamlit reruns your .py file from top to bottom
3. Updated widgets are sent back to the browser
No Flask routes or HTML templates — Python only. The rerun model matters when you add st.session_state in the next lesson.

Install

Activate the venv from Project Setup, then install Streamlit. See Streamlit installation for platform-specific notes.

pip install streamlit

Page layout

st.set_page_config, st.title("HTML Questions"), and st.chat_input — submit handling comes in Streamlit Chat Message History. First load shows the title and input bar (screenshot).

import streamlit as st

st.set_page_config(page_title="HTML Questions")
st.title("HTML Questions")

st.chat_input("Type a question…")
Introduction to Streamlit — HTML Questions at localhost:8501 with title and st.chat_input
First load at localhost:8501 — st.title("HTML Questions") and st.chat_input only. Submit handling is in Streamlit Chat Message History.

The demo script

introduction_to_streamlit_demo.py adds a launcher block so python introduction_to_streamlit_demo.py starts the server. Use streamlit run in day-to-day work.

◇introduction_to_streamlit_demo.py
"""introduction_to_streamlit_demo.py"""
# launcher when run with python
st.set_page_config(page_title="HTML Questions")
st.title("HTML Questions")
st.chat_input("Type a question…")

Download the code

introduction_to_streamlit_demo.py

Starter page — no LangChain yet

Download .py
Venv from Project Setup. Install Streamlit with pip install streamlit.

Run it

Activate the venv from Project Setup, then from the folder that contains the demo file:

streamlit run introduction_to_streamlit_demo.py
PowerShell — (.venv) active
(.venv) PS C:\projects\langchain-course> streamlit run introduction_to_streamlit_demo.py
You can now view your Streamlit app in your browser.
Local URL: http://localhost:8501
Paste http://localhost:8501 if the browser does not open. Keep the terminal open while the app runs.

Quick reference

  • streamlit run app.py — starts the local server; default URL http://localhost:8501.
  • st.session_state — dict that survives reruns in one browser tab (used in the next lesson).
  • st.set_page_config() — call once at the top for title and icon.
  • Install: pip install streamlit.
  • Docs: Streamlit docs · Installation · Create an app.

What's Next

Next: Streamlit Chat Message History — store turns in st.session_state and handle st.chat_input submits.

← PREVIOUS
MongoDB Chat Message History
NEXT →
Streamlit Chat Message History