Spreadsheet Encoding for LLMs

A Python package that implements encoding spreadsheets for Large Language Models.

96% Token Reduction
5x Faster Processing
100% Structure Preserved
Terminal
$ pip install sheetwise
$ sheetwise --demo
Created demo spreadsheet: (100, 30)
Compression ratio: 1.3x
Token reduction: 5.1x
✓ Ready for LLM processing

Key Features

Intelligent Compression

Up to 96% reduction in token usage while preserving semantic information and structure.

Structural Analysis

Identifies and preserves important structural elements in spreadsheets.

Multi-Table Support

Handles complex spreadsheets with multiple tables and preserves their relationships.

Structural Analysis

Identifies and preserves important structural elements like headers and data types.

Easy Integration

Simple API that works with pandas DataFrames and common spreadsheet formats.

Configurable

Customizable compression parameters to fit your specific use case and requirements.

Quick Start

Three simple steps to compress your spreadsheets

1

Install

pip install sheetwise
2

Import & Load Data

from sheetwise import SpreadsheetLLM import pandas as pd df = pd.read_csv("your_data.csv")
3

Compress & Get Results

sllm = SpreadsheetLLM() compressed = sllm.compress_and_encode_for_llm(df) stats = sllm.get_encoding_stats(df) print(f"Reduced by {stats['compression_ratio']:.1f}x")

Ready to get started?

Transform your spreadsheets for LLM processing in minutes