You can now translate your Zendesk Help Center from Lingo.dev.

The Zendesk Help Center integration connects your Help Center to a localization engine you host on Lingo.dev. Choose the categories, sections, and articles to cover, and Lingo.dev translates their titles, descriptions, and bodies into your target languages, then writes each translation back to Zendesk. Article bodies translate as HTML, so links, formatting, and embedded media survive the round trip.

There are two ways to keep your Help Center translations up to date. The first, Sync, translates only the articles you have added or edited since the last run. The second, Retranslate, rebuilds every article in scope from scratch, which is what you want after changing the localization engine's glossary, brand voice, or rules and needing your existing translations regenerated against the new configuration.

Human review can run on only the translations that need it.

Human review is a stage in the localization pipeline: a translation pauses so a person checks it before it ships. Until now, turning it on sent every string to a reviewer. Now it is selective. Set a quality threshold on the engine, and only translations that score below it are routed to a human. Everything at or above the threshold ships automatically. Reviewers spend their time on the translations most likely to need a fix, review cost drops, and high-confidence output no longer waits in a queue.

Full changelog: lingo.dev/changelog

Seed a whole glossary from the MCP server, across every locale at once.

Adding glossary terms through the MCP server no longer goes one term at a time. Ask your AI assistant to add a whole set of terms in a single step, spanning as many locales as you need, and it seeds the glossary in one pass. Building out a large termbase used to be slow, adding each term on its own call. Now it lands in one go, and the terms start steering translations right away.

Also shipped

• Docusaurus output paths in the CLI. Translated pages now land in the folder layout Docusaurus expects, so the localized site renders without moving files by hand.

• Fuller YAML frontmatter translation. Nested and list frontmatter fields, like SEO or FAQ metadata, are now translated too, not just simple top-level ones.

Localization Engine Suggestions now come from the corrections human reviewers make, not only from low AI-review scores.

When a reviewer edits a translation whether in-house or through an external provider the localization engine compares the change against what it first produced. When the same fix recurs across jobs, or a correction clearly sets a term, it proposes the glossary, instruction, or brand-voice edit behind it. Apply writes the change into the engine, and the next translation runs with the fix in place.

Suggestions can now be approved or dismissed in bulk select a set and act on all of them at once, instead of one card at a time.

Full changelog: lingo.dev/changelog

Audit logs are now available an append-only record of every state-changing action in your organization.

Each change to engines, glossaries, instructions, brand voices, AI reviewers, API keys, members, and roles is captured with the actor, action, target, and request context. The log lives in the dashboard and filters by user, action category, target, and date so you can see exactly who created, updated, or removed a glossary item three weeks ago, even if the actor is no longer on the team.

Also this week:

Glossaries, rulesets, and brand voices now live at the organization level and attach to any engine.

Build a glossary, a ruleset of linguistic rules, or a brand voice once, then attach it to any engine that needs it. Edit a shared asset and every engine it's attached to picks up the change. A new engine no longer starts from an empty configuration: attach what you already have and it inherits it.

Also this week:

• Take over a translation waiting on external review. When a job is parked on an external professional reviewer, an authorized member of your own team can open the pending review in the dashboard and submit the decision, so the job continues instead of waiting on the external round trip.

• A runnable example repo for every framework. Each is a real project with .lingo/config.json and translations committed clone it, run lingo link, and push: iOS, Android, Flutter, React + Vite, Next.js, Rails, Laravel, Markdown, Markdoc, OpenAPI, and TypeScript modules.

The async CLI now handles more file formats end to end: TypeScript locale modules, YAML, Gettext PO, Flutter ARB, Android strings, and Xcode .strings, String Catalogs (.xcstrings), and .stringsdict. Point a files[] entry at the source file and every lingo push runs each one through your engine's full pipeline. Translatable text is translated; everything structural is preserved keys, IDs, metadata, placeholders, and code.

Full changelog: https://lingo.dev/en/changelog

The localization pipeline wraps the core translate step with optional stages, each independently toggleable per engine and overridable per job:

• Pre-localization AI edit. An AI agent cleans the source payload before translation, so a single source error doesn't propagate across every target locale.

• Post-localization human edit. Sends the translation to a qualified human translator. The job pauses on a webhook until the edit returns.

• Post-localization AI review. Reconciles the human output against the engine's glossary, brand voice, and instructions.

• Back-translation check. Translates the final output back into the source locale and compares. The agent flags semantic drift by severity and auto-adjusts on major or critical drift.

Turn on what you need, leave the rest off.

AI reviews got more accurate: false positives dropped by 59.7%.

A false positive is a review that marks a correct translation as failing fewer of them means fewer alerts to check by hand, and review scores that can be trusted. We measured the drop on our internal benchmark. The biggest change: when a review doesn't apply to a given string, it no longer counts against the score, so scores reflect only the reviews that genuinely apply.

The reviewer setup form now shows example instructions as a placeholder for each review type.

Also this week: