Define the AI task and success criteria
Write the acceptance criteria before grading outputs.
AI Engineering tool
Prompt Evaluation Worksheet
Compare prompt versions using accepted results, retries, review time, and a fixed test set.
Runs in your browser with editable illustrative assumptions. It does not call a model, send entered content to a server, or produce a current market quote.
Interactive calculator
Prompt evaluation worksheet
min
Use this worksheet to compare prompt versions against the same cases. A score is only useful when the test set represents the work users actually send.
First-attempt acceptance80.0%
Retry rate25.0%
Review time for this set1.0 hours
Record prompt version, model, settings, tokens, latency, and grader notes beside these results. This worksheet does not call a model or store test content.
Starting values are illustrative defaults you can edit — not live ComputeTape Market benchmark prices. Replace them with a real quote.
How to use it
Make the assumption visible
Use the same representative cases for each prompt version. Quality is only one dimension; retries, tokens, latency, and review time are workload costs too.
- Change one assumption at a time when comparing designs.
- Record the prompt, workload, or permission version beside the result.
- Treat the output as a planning aid and validate it against representative work before making a production decision.