GitLab localization:
ship every feature already translated,
wording owned by your whole team

With Prismy, new strings are detected automatically on every MR. No manual trigger, no sync to run. Translations are generated in one click, using your product's glossary and brand voice, and the MR can merge with production-ready wording. There's no separate platform to log into, no back-and-forth before a release. Teams typically go from opening a MR to shipping a translated feature in the same development cycle, instead of waiting for a separate localization pass after the fact.

Adding a new language takes minutes in Prismy. Update your language settings and every existing MR workflow extends to cover it automatically. There's no new pipeline to configure, no new sync to maintain. Teams scaling from 2 to 10+ languages don't add headcount or process complexity.

Prismy connects to your GitLab repository as a native app and monitors every Merge Request. When a MR diff includes changes to translation files, new keys added, existing strings modified, Prismy automatically checks which translations are missing across your active languages and posts a comment on the MR with the details. No configuration required beyond the initial connection.

When a developer introduces new strings in a Merge Request, Prismy automatically attaches the MR context to each new key in its editor: the MR description, any screenshots attached to the MR, and any Figma design links referenced. When a translator opens a key, they see this context directly alongside the string, so they can translate with full understanding of the feature and its design intent.

Yes, specifically because it knows your product. Generic AI tools don't know your terminology, your brand voice, or the conventions your team has established. Prismy's AI uses your PM-configured custom instructions, your glossary, and the context from every translation your team has already approved. The output is consistently on-brand and correct in your specific vocabulary.

Yes. Prismy supports both GitLab.com and self-managed GitLab instances. If your organization runs GitLab on its own infrastructure for security or compliance reasons, Prismy can connect to it. For teams with stricter requirements, Prismy also offers a hosted mode where translation files are stored in Prismy's database rather than requiring a direct repository connection.

Yes, and that's a core part of the value. Once the developer installs Prismy, PMs and translators can make wording changes directly, on feature branches (changes commit immediately) or in production (Prismy opens a conflict-free MR automatically). They don't need GitLab access or JSON files. Reviewers can also use Prismy's web app or Chrome extension to edit copy directly in the live product, WYSIWYG-style.

Phrase and Lokalise both use a file sync model: they pull your translation files into their platform, you translate there, and they push back. Prismy lives inside your Merge Requests: it detects missing translations in every MR diff, provides translators with full MR context (description, screenshots, Figma links), and generates AI translations using your specific product terminology and voice. No separate sync to configure or maintain.

Prismy works at the MR level, so it doesn't require changes to your CI/CD setup. If you want to block deployments when translations are missing, you can add Prismy's GitLab check to your pipeline. It blocks MR merges until required translations are approved. Configurable via the API and CLI.

Prismy is stack-agnostic and supports all common i18n formats: JSON, YAML, PO/POT, XLIFF, Android strings XML, iOS .strings, and more. It works with any framework or i18n library that stores translation files in your repository.