Python prompt compression guide

Installation

pip install supercompress

That is it. The package includes the compression policy, Python bindings, and CLI tool. No GPU, no model download, no external dependencies.

Basic usage

from supercompress import Compressor

comp = Compressor()
result = comp.compress(
    context="""Alice has been a Premium subscriber since 2021.
Her account type is Enterprise.
She last logged in on June 14, 2026.
Her current plan costs $199/month.""",
    query="What plan is Alice on?"
)

print(result.compressed_text)
# Output keeps: "Her current plan costs $199/month."
# Drops: login date, subscriber since info (irrelevant to the question)
print(f"Saved {result.tokens_removed} tokens")

Integration with OpenAI

import openai
from supercompress import Compressor

comp = Compressor()
client = openai.OpenAI()

def ask_with_compression(context, query):
    result = comp.compress(context, query)
    response = client.chat.completions.create(
        model="gpt-4o",
        messages=[
            {"role": "system", "content": "Answer based on the context."},
            {"role": "user", "content": result.compressed_text},
        ]
    )
    return response.choices[0].message.content

Frequently asked questions

Does SuperCompress work with Python 3.9+?

Yes. SuperCompress supports Python 3.9 through 3.13.

Can I use it with async Python?

Yes. The compressor is synchronous and fast (~60ms), so it works in async contexts without blocking.