streamlit-aggrid-v2

AG Grid component for Streamlit - interactive tables, editing, filtering, sorting, grouping & more.

Built on AG Grid v34.3.1 with Streamlit Custom Components v2.

Acknowledgments

This project is a continuation of streamlit-aggrid, originally created by Pablo Fonseca. Huge thanks to Pablo for building and maintaining the first version that the Streamlit community relied on for years.

Install

pip install streamlit-aggrid-v2

Note: The Python import stays st_aggrid - only the pip install name changed.

Quick Start

from st_aggrid import AgGrid
import pandas as pd

df = pd.read_csv(
    "https://raw.githubusercontent.com/fivethirtyeight/data/master/airline-safety/airline-safety.csv"
)
AgGrid(df)

streamlit run your_app.py

Why v2?

The original streamlit-aggrid is built on Streamlit's legacy iframe-based component model (CCv1). Migrating to CCv2 is a significant architectural change - this rewrite does that while keeping the same Python API.

	streamlit-aggrid (original)	streamlit-aggrid-v2
Component model	CCv1 - iframe + postMessage	CCv2 - direct DOM rendering
AG Grid version	v31	v34.3.1
Themes	3 themes, no dark mode for quartz	4 themes, automatic dark/light detection
Build toolchain	webpack / CRA	Vite + ESM
React	React 17	React 18
Known bugs	25 open issues in codebase	All 25 fixed
Event listener cleanup	Leaks on unmount	Proper cleanup via componentWillUnmount
State management	Direct this.state.api mutation	gridApiRef pattern (React-safe)
Style/script injection	Duplicates on re-render	Deduplicated
Mutable defaults	grid_response={}, update_on=[]	Fixed with None + inline init
Path handling	Wrong variable, missing method	Corrected
Custom themes	Not supported	StAggridTheme with color schemes, icon sets, params
Maintenance	Inactive since 2023	Active, semantic versioning, CI/CD
Python API	from st_aggrid import AgGrid	Same - fully backward compatible

Migration: Change pip install streamlit-aggrid to pip install streamlit-aggrid-v2. No code changes needed.

Features

• No iframes - AG Grid renders directly in the Streamlit DOM via CCv2, eliminating postMessage overhead.
• Theming - All four AG Grid themes (quartz, alpine, balham, material) with automatic dark/light mode detection. Custom themes via StAggridTheme with color schemes, icon sets, and param overrides.
• Editing - Cell editing with singleClickEdit, value parsers, and change detection.
• Row selection - Single, multiple, and checkbox selection with pre-selected rows.
• Filtering & sorting - Column filters, floating filters, quick search, and multi-column sort.
• Column configuration - Pinning, resizing, reordering, auto-sizing, and column groups via GridOptionsBuilder.
• Cell renderers - Custom cell rendering with JsCode (stars, badges, progress bars, buttons).
• Row styling - Conditional row/cell styling via getRowStyle, rowClassRules, and cellStyle.
• Enterprise features - Row grouping, pivot mode, status bar, side bar, Excel export, cell selection, sparklines (requires AG Grid license).
• Data return modes - AS_INPUT, FILTERED, FILTERED_AND_SORTED, MINIMAL, and CUSTOM via the collector pattern.
• Modern build - Vite + ESM replaces the legacy webpack/CRA toolchain.
• Bug fixes - 25 bugs fixed from the original codebase.
• Backward compatible - Existing AgGrid(), GridOptionsBuilder, JsCode code works with just a pip install change.

API Overview

from st_aggrid import AgGrid, GridOptionsBuilder, JsCode
from st_aggrid import GridUpdateMode, DataReturnMode, ColumnsAutoSizeMode

# Build grid options from a DataFrame
gb = GridOptionsBuilder.from_dataframe(df)
gb.configure_pagination(paginationAutoPageSize=True)
gb.configure_selection("multiple", use_checkbox=True)
gb.configure_default_column(filter=True, sortable=True, floatingFilter=True)
grid_options = gb.build()

# Render the grid
response = AgGrid(
    df,
    gridOptions=grid_options,
    height=400,
    theme="quartz",                          # or "alpine", "balham", "material"
    update_mode=GridUpdateMode.MODEL_CHANGED,
    data_return_mode=DataReturnMode.FILTERED_AND_SORTED,
    allow_unsafe_jscode=True,                # required for JsCode renderers
)

# Access results
selected_rows = response.selected_rows
filtered_data = response.data

Live Demo

Check out the full showcase with 12 interactive examples:

st-aggrid.streamlit.app

License

MIT