Skip to content
Locize
LoginRegister

Blog

Product updates, guides, and customer stories.

Do you still need a TMS when your coding agent translates the app?Guides

Do you still need a TMS when your coding agent translates the app?

Many teams now translate with a coding-agent routine: a Claude Code or Codex skill that drafts the missing strings on a schedule and opens a pull request. When that routine is enough, what it has to rebuild by the tenth release, and three ways to keep it and put a memory, a reviewer and a record under it, from one added command in CI to a routine that reads the glossary before it drafts.

November 3, 2026• 13 min read
AI translations that respect your i18n formatProduct

AI translations that respect your i18n format

Locize now checks every AI and machine translation against its source before storing it: the syntax, the placeholders, the tags and the plural forms the target language needs. The AI prompts carry the rules of MessageFormat (often called ICU) and Fluent, issue weights and "fix with AI" sort what is left, and MessageFormat 2 is a project format.

October 21, 2026• 8 min read
ICU and MessageFormat 2 in i18next: typed arguments, linting and the App RouterGuides

ICU and MessageFormat 2 in i18next: typed arguments, linting and the App Router

i18next has supported ICU MessageFormat through i18next-icu for years, and Unicode MessageFormat 2 through i18next-mf2. In October 2026 both became first-class options next to the i18next format: debug warnings when syntaxes are mixed, type-checked ICU and MF2 arguments, an i18next-cli mode that extracts one key per message and lints plural categories per locale, IcuTrans in React Server Components, and next-i18next providers for the App Router. What changed, with code that runs.

October 18, 2026• 9 min read
One product, many customers: per-customer translations with tenants, a worked exampleGuides

One product, many customers: per-customer translations with tenants, a worked example

White-label and B2B products often need the same app with each customer's own wording: "client" instead of "patient", a different cancellation policy, the customer's name instead of yours. The ways teams handle it (copies, custom locale codes, namespaces, interpolation, tenant projects), and a real run with one base project and three customer tenants: overrides, what happens when the base changes, how the app loads the right customer, and what to watch.

October 14, 2026• 8 min read
How to choose a TMS in 2026: pricing models explained, one project priced at nine vendorsGuides

How to choose a TMS in 2026: pricing models explained, one project priced at nine vendors

Translation management systems meter different things: words in all languages, words in target languages, processed words, source keys, strings or seats. We priced one reference project (2,000 keys, 6 people, 3, 8 or 15 target languages) at nine TMS vendors, Locize included, from their public pricing pages on October 13, 2026. What each pricing model rewards, where it gets expensive, and what the list price leaves out.

October 13, 2026• 13 min read
A coding agent localizes a feature: confidence scores, review and the record, step by stepGuides

A coding agent localizes a feature: confidence scores, review and the record, step by step

A real run, not a diagram: Claude Code adds a snooze feature to a React app, reports eleven keys to Locize over MCP, and gets them back in six languages within seconds, each with a confidence score. What the scores caught, what they missed, what the agent fixed by itself, what the review still changed, and what the provenance export records about each string.

October 12, 2026• 9 min read
Update Translations Without Redeploying: Runtime Loading, Caching, and Why Users See Stale StringsGuides

Update Translations Without Redeploying: Runtime Loading, Caching, and Why Users See Stale Strings

How to change the text of a web or mobile app without shipping a new release: load translations at runtime instead of bundling them, and understand the caches (browser, service worker, localStorage, CDN) that decide when users actually see the change. Covers i18next backends, Cache-Control, the usual reasons translations do not update after a deploy, and when bundling is still the right call.

October 11, 2026• 11 min read
Git-synced translations vs runtime delivery: which one fits your CI/CD pipeline?Guides

Git-synced translations vs runtime delivery: which one fits your CI/CD pipeline?

Translation files kept in git and synced through CI, translations loaded from a CDN at runtime, or both? How each architecture works end to end, an honest decision table (who edits, release cadence, SSR and offline, pull-request review, rollback, CDN cost, confidential strings, native mobile), how teams run it today, and what neither approach solves.

October 10, 2026• 10 min read
The EU AI Act transparency Code of Practice: who signed, and what it means for translated contentGuides

The EU AI Act transparency Code of Practice: who signed, and what it means for translated content

As of September 9, 2026, the European Commission's Code of Practice on transparency of AI-generated content lists 83 provider and 152 deployer signatories, and the list keeps growing. What the Code is, what signing commits a company to, why not signing is not non-compliance, and what it changes, and does not change, for machine- and AI-translated websites and apps under Article 50.

September 9, 2026• 10 min read