Type your text, then translate to japanese _ (or any language) and OpenCues rewrites the text in place. The instruction phrase is consumed, your translated text remains, and you can cycle back to the original at any time.
Three translation shapes
- Whole draft:
<your paragraph> translate to spanish _rewrites everything before the instruction. - Quick lookup:
the word for love in japanese _answers in place via the fluid blank; only the_is replaced, your question stays put. - Template fills:
hello world in german is _keeps your sentence and fills only the blank.
Why it beats the copy-paste loop
The usual path is: leave the field, open a translator, paste, copy, return. OpenCues collapses that to one keystroke because the request lives inside the text itself. If the model cannot resolve it, the _ just stays; nothing is lost.
Substitutions are reversible: navigate to the result with Ctrl+Alt+←/→ and press Ctrl+Alt+↓ to restore the original text. A lookup only ever replaces its own short span, so the rest of your draft is untouched.
Related: What is a blank? · AI in any text field · Inline AI for non-native English writers · Supported languages · The dictionary blank · GitHub repo