Prompt Diff
Compare the words, the whitespace and the cost assumptions. Keep the full text in view, and judge answer quality with a separate test.
Model data. Verified-price snapshots from . Details
35 of 39 tracked models have verified token rates in the saved catalogue. Rates come from the OpenRouter public API. Models missing from that catalogue have unknown pricing and are excluded from cost estimates and chat selection.
Rates are USD per million text tokens. Caching, reasoning tokens, images, tool charges, provider routing and long-context tiers can change the bill. Check the provider before budgeting.
Context limits have their own dated review. The Context Budget Planner shows the catalogue and top-provider quotes separately, using the smaller known limit. Output caps are a separate constraint. Weight links identify repository revisions whose file lists were checked. Licence labels come from the model cards and are not an open-source certification. Check the actual terms and access conditions. Published weights do not make a hosted API or self-hosting free.
TRY A CHANGE
More precise? Shorter? Just different?
Try a pair, inspect the changes, then test the output separately. A smaller prompt is not automatically a better prompt.
Compare saved Workbench versions
Read saved prompts from this browser. Loading here does not change the saved library.
0 characters
0 characters
Put the size beside the price.
| Measure | A | B | Delta |
|---|---|---|---|
| Characters | 0 | 0 | 0 |
| Estimated tokens | ~0 | ~0 | 0 |
| Input cost, USD | $0 | $0 | $0 |
Token counts use a text heuristic. There is no measured accuracy range. Estimates exclude message overhead, output, reasoning and caching. Price and length do not measure answer quality.
How the comparison works
Compare up to 200,000 characters per editor. Words are runs of non-whitespace text, and spaces, tabs and line breaks are preserved. Character totals count Unicode code points. A word comparison does not establish whether the meaning or the model's answer will stay the same.
A separate worker finds shared words and whitespace. Detailed comparison is limited to two million table cells and 600 display segments. Larger differences use explicitly labelled blocks. The worker is stopped after two seconds, and editing either prompt clears the previous result.
Build and save a prompt in the Workbench ↗ · Inspect the comparison code ↗