Module Profile
Net-Base AI Multilanguage at a glance
Net-Base AI Multilanguage is not a superficial language switcher for us, but a technical layer for base language, target languages, browser detection, queue operation, protected terms, and editorial preview. This is exactly what keeps language controllable, even when many pages, topic areas, legal texts, and SEO fields are connected.
German remains the technical lead
New content is created from an editorially maintained German baseline and is transferred from there into active target languages in a controlled manner.
Company and product names remain stable
Protected terms ensure that brands, project names, and company-specific terms do not fragment through automatic translation.
Queue instead of restless individual updates
Changes to pages, menus, taxonomies, theme texts, and SEO fields are collected, propagated, and visibly tracked in the backend.
What this module delivers in concrete terms
Manages the base language German, active target languages, browser detection, translation queue, and backend previews for language variants.
Scope
- The base language is always German; all other active languages are generated from it.
- Supports pages, posts, public custom post types, menu items, taxonomies, website subtitle, as well as fixed theme and plugin texts.
- Automatically re-translates changed content when automation is active.
- Displays language variants in the frontend via browser language or ?lang=xx.
- Protected terms such as company, product, and project names remain unchanged.
Language control in the frontend
- Browser detection applies only to activated target languages.
- ?lang=en forces a language directly, ?lang=de switches to the base language, ?lang=auto resets to automatic detection.
- The current language selection is kept via the plugin cookie and is evaluated again on subsequent page views.
- If nothing exists for a language yet, output falls back to German in a controlled manner.
Queue and operations
- New or outdated texts end up in a translation queue.
- The queue is processed in the background via cron and an additional kickoff.
- The number of parallel workers is configurable; additional workers run as separate signed queue processes with database claiming instead of as unsafe PHP threads.
- Batch size and interval control speed, API load, and cost.
- Failed jobs can be restarted in bulk.
Backend helpers
- In the editor, there are language-variant tabs for preview per active language.
- There you can see per field whether the translation is current, pending, outdated, or failed.
- The admin page shows overall status, last errors, and the current queue operation.
- When saving the settings, the entire inventory can be queued again.
- Default protected terms are visible in a dedicated tab; custom protected terms can be maintained there as an additional list.
- Protected terms are masked before the AI request and then re-inserted exactly afterward so that brand and product names are not transliterated.
Interaction with SEO
- The language module provides the frontend language on which hreflang, canonical, and SEO metadata build.
- New active languages can be translated along by the other Net-Base plugins, provided they register their content.
- SEO titles and descriptions are not maintained here, but are connected linguistically by the SEO modules with this system.
Limits and notes
- Without a valid API key, new jobs remain pending or fail.
- Very many active languages increase queue runtime, cost, and the amount of content that needs follow-up maintenance.
- The base language remains editorially leading. Target languages should still be checked on a sample basis.
Classification
Multilingualism is the website’s technical language layer. It does not provide an SEO strategy and does not replace editorial control in target languages.
Why this layer is important for international visibility
As soon as website, magazine, and SEO become multilingual, a simple language switcher is no longer sufficient. Then it is about language-specific canonicals, consistent naming, clean fallbacks, and the question of how changed content can technically follow through into multiple languages in a calm, controlled way.
Don’t treat languages as a side issue
Multilingualism only becomes economically interesting when language, SEO, protected terms, and editorial work are technically thought through together. This module is built for exactly that.