WoluTools
Find a tool

Localization Package QA

Localization Package QA Online: catch localization defects before release.

Turn a localization handoff into a precise unit-level correction list.

START HEREChoose documentsXLF, XLIFF, SDLXLIFF, TMX, PO, TS
Choose documents Try prepared example
Maximum
25 files · 64 MB total
File handling
Draft 2 hours · encrypted result 24 hours
FREE3 jobs a dayFiles up to 10 MB
PROUp to 200 jobs a day€12.99/month or €89.99/year
See plans

Larger files need Pro. Failed jobs are never counted.

  • Prepared example is free
  • 3 free jobs a day
  • Cancel anytime
EXAMPLE RESULTSample data
Units · 386
14
92%
386
Targets present · 92%
14 release-blocking units need correctionHTML, protected localization summary and unit CSV files, technical JSON and manifest
Units
386
Blocking findings
14
Targets present
92%
Formats, limits & file handling
Works with
XLF, XLIFF, SDLXLIFF, TMX, PO, TS
Limit
25 files · 64 MB total
You receive
HTML, protected localization summary and unit CSV files, technical JSON and manifest
File handling
Draft 2 hours · encrypted result 24 hours

Turn a localization handoff into a precise unit-level correction list.

THREE CLEAR STEPS

From your documents to a usable result.

  1. 01
    Add your input

    Limits and supported formats are visible before you begin.

  2. 02
    Review the result

    Check the preview, findings or artwork before you download.

  3. 03
    Download

    Use your free job, plan or credits. Failed jobs are never counted.

Prepared result preview

See the decision before sharing a real file.

Turn a localization handoff into a precise unit-level correction list. This prepared example uses fictional data and the same evidence structure as the server export.

✓
Nothing is uploadedPublic demo only

Localization QA

14 release-blocking units need correction

Every finding points to the file, unit and deterministic rule; wording is never generated or translated.

Actual prepared coverage: 6 files · 386 units · XLIFF, PO and Qt TS

  • ErrorPlaceholder mismatchcheckout.xlf · unit 184 · {price} is missing in the target
  • ErrorMissing targetdesktop.ts · unit settings.notifications · target is empty
  • ReviewLinguistic meaningWording quality still needs a human reviewer
Units
386
Blocking findings
14
Targets present
92%
Show the verified package

HTML, protected localization summary and unit CSV files, technical JSON and manifest

Engine localization-qa-v2 · rules localization-standard-profile-auth-2026-08-31

One focused workflow

From bounded input to evidence you can review

  1. Add the source.Accepted inputs are XLF, XLIFF, SDLXLIFF, TMX, PO, TS. The visible limit is 25 files · 64 MB total.
  2. Confirm the job.Every assigned source number, selected filename and setting stays visible before the confirmed SHA-256 input hash is accepted. Product-specific mappings remain explicit settings rather than guessed roles.
  3. Process in isolation.A one-job networkless sandbox receives only this job and cannot access accounts, queues, encryption keys or other customer storage.
  4. Review exact findings.Check XLIFF, TMX, PO and Qt TS packages for broken placeholders, tags, numbers, whitespace and missing targets without sending translations away.
  5. Download the evidence.HTML, protected localization summary and unit CSV files, technical JSON and manifest Every output hash is covered by the coordinator-created manifest.

What this product does not claim

Every finding points to the file, unit and deterministic rule; wording is never generated or translated.

Clear answers

Localization Package QA FAQ

Which handoff formats are supported?

XLF, XLIFF, SDLXLIFF, TMX, PO and Qt TS, up to 25 files and 64 MB per job.

What counts as a release-blocking finding?

A defect that would break the build or the running interface, such as a placeholder present in the source and missing from the target, broken inline tags, or an empty target where one is required.

Does the tool judge whether a translation is correct?

No. The rules are deterministic and structural. Meaning, tone and register still need a linguist, and no wording is ever generated or machine translated here.

How precise are the findings?

Each finding names the file and the translation unit identifier, so a translator opens the exact segment instead of searching the package for it.

What do I get back?

An HTML report, a protected localization summary CSV, a unit-level CSV, technical JSON and a manifest covering the outputs.

Available now

Catch localization defects before release.

Inspect the fictional result now. Your own files run here without an account: 3 free jobs a day, or up to 200 a day with Pro.

Inspect prepared resultFind another tool