Dify prompt compression

Adding as a custom tool

In Dify, create a custom code node that calls SuperCompress before passing context to your LLM node:

import requests

def main(context: str, query: str) -> dict:
    resp = requests.post(
        "https://supercompress.dev/api/v1/compress",
        headers={"X-API-Key": "YOUR_KEY"},
        json={"context": context, "query": query}
    )
    data = resp.json()
    return {
        "compressed": data["compressed_text"],
        "savings": data.get("original_tokens", 0) - data.get("kept_tokens", 0)
    }

Frequently asked questions

Can I use SuperCompress in Dify workflows?

Yes. Add it as a code node between your knowledge retrieval and LLM nodes.

Does it require API configuration in Dify?

Yes. Add your SuperCompress API key as an environment variable in Dify.