What this tool does
Evaluation Rubric Generator drafts weighted scoring criteria with level descriptors and machine-readable JSON.
← All Tools
📝 Evaluation Rubric Generator
Build weighted scoring rubrics with clear criteria, levels, and copyable JSON or Markdown.
Why it is useful
It helps create clearer judgment standards before comparing answers, submissions, or generated content.
How it works
The browser turns task goals and criteria into a 1-5 scoring rubric.
Best input
Describe what success looks like, common failure modes, and must-have constraints.
Review tip
Test the rubric on a few examples and revise unclear criteria before using it broadly.
Privacy note
Rubric drafting happens locally in your browser.
Important limitation
Generated rubrics need human review. Vague criteria can produce inconsistent scoring even when the format looks complete.
Quick answers
Can this grade work automatically?
No. It drafts criteria; scoring still needs a grader or evaluator.
What makes a good rubric?
Clear dimensions, observable criteria, and examples of strong and weak performance.
Common Use Cases
Create rubrics for LLM output evaluation
Generate structured rubrics to consistently evaluate model outputs.
Standardize evaluation criteria
Define clear scoring dimensions when comparing prompts or models.