AI tools

Free AI Rubric Generator

Turn your assignment goals into a draft rubric with observable criteria and clear level descriptors. No points or standards added for you.

How it works

  1. Paste your assignment prompt and the goals students must show, such as clear setup, accurate measurements, or a supported claim.
  2. Choose three or four achievement levels. Name your assessment criteria in the text box; the draft uses at most four.
  3. Review each criterion heading and its descriptor for your selected levels. Edit any wording that is vague, unfair, or not observable in student work.

Examples

Sample results for the inputs shown, so you can see the format and quality before you try it.

Example 1: input

Assignment and criteria: Assignment: explain how a paper bridge was tested. Criteria: clear setup, accurate measurements, acknowledge limitations. Use three achievement levels. No numerical scoring.

Output style: Three levels

Example 1: output
Criterion: Clear Setup Developing: Describes some materials or steps, but the sequence is incomplete, out of order, or missing one or more key parts of how the bridge was built or tested. Meets: Describes the bridge materials and the testing steps in order, so a reader could follow how the test was carried out. Extends: Describes the setup precisely and concisely, including the order of steps, how the load was applied, and any conditions that could affect the test. Criterion: Accurate Measurements Developing: Reports one or more measurements that are missing units, inconsistent with each other, or not linked to what was measured. Meets: Reports measurements with units and connects each measurement to the part of the bridge or test it describes. Extends: Reports measurements with units, shows how they were obtained, and notes where variation or uncertainty in the measurements could occur. Criterion: Acknowledges Limitations Developing: Mentions a limitation only in general terms, or states a limitation without explaining how it affected the test. Meets: Names at least one specific limitation of the test and explains how it could affect the results. Extends: Names specific limitations, explains their likely effect on the results, and suggests how the test could be changed to reduce them.
Example 2: input

Assignment and criteria: Assignment: write a book review with a claim, textual evidence and an explanation of the evidence. Use three achievement levels. No spelling criterion requested.

Output style: Four levels

Example 2: output
Criterion 1: Claim - Beginning: States a topic or opinion about the book without a position the review will defend. - Developing: States a claim about the book, but the claim is vague or only loosely tied to the rest of the review. - Meets: States a clear claim about the book and keeps the review focused on supporting it. - Extends: States a clear, specific claim and sustains it throughout, including in the conclusion. Criterion 2: Textual Evidence - Beginning: Refers to the book in general terms without pointing to specific passages or events. - Developing: Includes evidence from the book, but it is brief, unclear, or only partly relevant to the claim. - Meets: Includes relevant evidence from the book, such as quoted lines, paraphrased passages, or specific scenes, that supports the claim. - Extends: Includes well-chosen evidence from multiple points in the book that directly and consistently supports the claim. Criterion 3: Explanation of Evidence - Beginning: Presents evidence without explaining how it connects to the claim. - Developing: Explains the evidence, but the connection to the claim is general or incomplete. - Meets: Explains how each piece of evidence supports the claim. - Extends: Explains how each piece of evidence supports the claim and addresses how the evidence fits the book as a whole or anticipates an alternative reading. Missing information: No grade level, word count, citation format, or required number of evidence items was supplied; none has been assumed.
Example 3: input

Assignment and criteria: Assignment: group poster about a local tree. Assess labelled parts, factual accuracy against class notes and legible layout. Use three levels. Do not score artistic talent.

Output style: Three levels

Example 3: output
**Labelled Parts** - Developing: Some tree parts are labelled; labels may be missing, unclear, or placed away from the part they name. - Meets: The main tree parts are labelled, and each label sits close enough to its part to be read without confusion. - Extends: All labelled parts are accurate and clearly placed, and any additional parts shown are labelled consistently with the rest. **Factual Accuracy Against Class Notes** - Developing: Several statements about the tree differ from the class notes or are missing where the notes give information. - Meets: The poster's statements about the tree match the class notes; any disagreements with the notes are absent. - Extends: Statements match the class notes and are specific where the notes are specific, with nothing added that the notes do not support. **Legible Layout** - Developing: Writing or labels are hard to read in places, or the arrangement makes parts of the poster difficult to follow. - Meets: Writing and labels are readable, and the arrangement lets a reader move through the poster in an order that makes sense. - Extends: Writing and labels are readable throughout, and the arrangement makes the tree's parts and the group's information easy to find quickly. Note: No points or grade levels are supplied, so none are included. "Artistic talent" is not assessed; layout is judged only for legibility and order.

Tips for better results

  • Write each criterion as something you can point to in the work, such as labelled parts on a poster, not effort or neatness.
  • If a goal is hard to observe, split it into smaller parts before generating, for example separate setup, measurements, and limitations.
  • Keep the same three or four criteria across levels so the only thing that changes is how well the work meets each one.
  • Read descriptors aloud and ask whether two reviewers would mark the same sample the same way. Rewrite any descriptor that fails this check.

FAQ

Is this free?

Yes. Email signup is required and also subscribes you to Something Big, our free newsletter. Daily usage limits apply. You can unsubscribe anytime.

Does it invent standards or points?

The instructions tell it not to. It only builds criteria and descriptors from the goals you supply. It does not add official standards, grade levels, alignment claims, or scoring. If you want points, you add them yourself. Check the draft against your source; generation can still make mistakes.

How many criteria and levels can it produce?

At most four criteria. Levels are three by default, or four if you select that option. Each descriptor is written to be specific and observable, but you must review and edit them to fit your assignment.

Who should review the draft?

A human must review it. The tool can miss context, produce wording that is too broad, or suggest descriptors that do not match your teaching. Check every level with real student work before you use it for grading.

More free AI tools

Related prompts

Prefer to use ChatGPT or Claude directly? These free prompts do similar jobs.

Get the best free AI tools and prompts every week

Something Big is a free AI newsletter read by 50,000+ professionals. About three short emails a week with the AI tools and prompts that actually work, plus what changed in AI and what to do about it.