Skip to content
PromptifyLab

Prompt Compare

Two versions of a prompt, side by side. It shows a word-level diff of what changed, a table of which structural ingredients each version has, and whether the new one is genuinely better structured or just longer. It also writes the prompt that makes a model judge them properly.

What changed

Green is added in B, red is removed from A.

Side by side

IngredientAB

Or have a model judge them


    

  


  
Nothing leaves your browser. Both versions stay on this device.







How it works


    

  • The diff is word-level, so a reworded sentence shows as a few changed words rather than a whole changed block.
    • 

  • Fourteen ingredients are checked in both: role, context, audience, format, length, tone, constraints, criteria, examples, reasoning, clarification, variables, source anchoring, structure.
    • 

  • It tells you what B added, what B dropped, and what the change costs in tokens per call.
    • 

  • The judge prompt runs both on the same input and forbids manufacturing a winner.
    • 



What it does not do


Every tool on this site says this plainly, because a tool that pretends to do more than it does costs you more than it saves.


    

  • This is a structural check, not a quality judgement. Structure correlates with quality; it does not guarantee it.
    • 

  • The only real test is running both on the same inputs several times each and comparing outputs.
    • 

  • Very long prompts fall back from word-level diffing.
    • 





Nothing you type here leaves your browser


There is no account, no server and no logging. The whole tool is JavaScript running on your own machine — close the tab and it is gone. You can check: open your browser’s network tab and use it.




Questions people ask




How do I know which prompt is actually better?
Run both on the same five inputs, three times each, and score against a rubric you wrote before looking. Anything less is an impression.




Why does my longer prompt perform worse?
Instruction compliance degrades as instructions accumulate, and long prompts often contain contradictions added over time. The ingredient table will show you if you dropped something while adding.







    

  • 

    Techniques
    

    Few-shot, chain of thought, role prompting — what holds up and what has been superseded.
    

        7 prompts
  
    • 

  • 

    Model Guides
    

    Where the models genuinely differ, and where the differences are overstated.
    

        7 prompts
  
    • 



Free, no sign-up. Last updated 18 September 2026.