<aside> 🚨 Want more notes like this? Subscribe to weekly strategy emails: seonotebook.com

</aside>

Hey Notebooker! Prompt Compression helps you take long prompts and compresses the token size so that you use fewer tokens for effectively the same prompt. This allows you to add more features to your CustomGPT Bots and allows any automations to run cheaper because they use fewer tokens.

I encourage you to experiment with different compression ratios and test your outputs, but once you’re happy with them. You will be able to do more with your CustomGPTs (because it enables you to get around the 8000 character limit), it’ll even save your time because your prompts will run faster AND it’ll save you money because you’re using fewer tokens! What’s not to love?!

Let’s get into the note! It’s pretty straightforward:

Step 1, watch this video:

https://www.youtube.com/embed/2jRDJE9jUWo

Step 2, try the Prompt Compressor for yourself:

SEO Notebook Prompt Compressor

Step 3, (for the hardcore people), install your own instance of LLM Lingua

https://github.com/microsoft/LLMLingua

And for those of you who are curious about the CustomGPT I demo’d in the video, here it is: