streamlit-aggrid-v2

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

Built on AG Grid v36.2.0 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.

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 and automatic font matching to Streamlit's theme (--st-font). 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 13 interactive examples:

st-aggrid.streamlit.app

AI Skills

A ready-to-use Claude Code / Claude Agent SDK skill lives at skills/streamlit-aggrid-v2/SKILL.md. Copy that folder into your project's .claude/skills/ (or the equivalent for your AI tool) and your copilot will know how to build grids with GridOptionsBuilder, wire up JsCode renderers, pick the right DataReturnMode, theme the grid, build tree-data views, and avoid common gotchas, without re-reading the whole repo every conversation.

License

MIT