Same lesson count.
Different critical token?
Compare already translated, user-aligned segments before a human language review. Exact token differences stay visible; matching tokens prove only structure, never semantic equivalence, translation quality or readiness to publish.
Source → target structural preflight
| Segment | Source preview | Target preview / language | Mechanical result | Difference types | Reviewer / decision |
|---|
Holds and human review queue
“Token match only” means the narrow extracted token bags match. It says nothing about omitted prose, mistranslation, tone, reading level, cultural fit, accessibility, legal effect or education equity.
Localization boundary: all free comparison stays in this browser. The paid request receives only the authorized, redacted segments you explicitly submit. Do not enter student, learner, client, employee, patient, account or credential data; private URLs; assessments; grades; accommodation records; or unpublished sensitive material. The tool does not translate, align sentences, evaluate meaning, detect every token, validate localized number/date conventions, certify accessibility or compliance, make education decisions, or publish. A qualified language reviewer and responsible content owner control every release.
$3 Structural Review Pack
Unlock an editable review pack for up to 500 user-aligned segments: token matrix, difference queue, language-review handoff, exception register, release sign-off and second-release usage and outcome review ledger.
- source-to-target token matrix
- number, date, URL, placeholder and markup difference queue
- qualified language-review handoff
- exception and owner register
- release sign-off and second-release usage and outcome review ledger
Enterprise substitutes and the deliberately narrower wedge
先核对当前来源与工具边界Check current sources and tool limits first