Skip to content
Aristo
Start for free

Prompt literacy for everyone

Write prompts that get the result you meant

Aristo grades your prompt against a published rubric, shows you exactly what is missing, helps you fix it yourself, tests the result on a real model and tracks your progress. In 15 languages.

Grade a prompt for freeSee the rubric

No account needed for your first analysis. Free accounts get five analyses a day.

How Aristo works

  1. 1

    Paste your prompt

    Any task, any language. Choose the model family you use, or let Aristo detect the task type.

  2. 2

    Get a graded report

    Nine criteria, each scored 0 to 4 against written anchors, with quotes from your prompt as evidence and one concrete next step per criterion.

  3. 3

    Fix it with the Professor

    Professor Aristo asks up to three questions about what is missing. Your answers become your prompt, in your own words.

  4. 4

    Optimise and test

    Ask for a rewrite tuned to Claude, GPT, Gemini or open models, see every change explained, then run both versions on a real model and compare.

Why a rubric, not a magic rewrite

Requirements first

A randomised trial found that learning to state requirements nearly doubled novice prompt quality, while generic prompt tips barely moved it. Aristo's rubric grades goal, context, audience and constraints before style.

Anchored, repeatable scores

Every criterion has written anchors, and each analysis takes the median of three independent judge samples, so the score means the same thing tomorrow.

You keep your voice

Coach mode completes your prompt with your own answers. Rewrites are shown as a diff with a reason for every change, and only called better when a blind comparison agrees.

The nine criteria

Public, versioned and the same for everyone. (rubric-2.0)

  • Requirements
    Goal and success criteria

    Is the desired outcome stated, and how will we know it is good?

  • Requirements
    Context and inputs

    Does the model have the facts, materials, data and situation it needs, clearly delimited?

  • Requirements
    Audience and purpose of use

    Who is the output for, and what will they do with it?

  • Requirements
    Constraints and boundaries

    Scope, must and must-not, length, sources, safety, what to do when uncertain.

  • Execution
    Task instructions

    Clear, ordered, unambiguous steps; one task or explicitly sequenced tasks.

  • Execution
    Output contract

    Format, structure, schema, length, language and tone of the result.

  • Execution
    Examples and references

    Worked examples or reference material where they genuinely help; varied, not over-fitted.

  • Craft
    Role and voice

    A perspective or expertise that changes the output; tone guidance.

  • Craft
    Model fit and hygiene

    Suits the target model family; no contradictions or shouting; long material before the question; no obsolete tricks.

Read the full rubric

Fifteen languages, one standard

Feedback is always written in the language you choose. The rubric is identical in every language, so a score in Portuguese means the same as a score in Japanese.

Ready to practise?

Grade your first prompt in under a minute.

Start for free

Aristo is new. There are no invented numbers on this page: usage figures appear only once there are real ones.