Panel rating · 6 judges · How to read the stars

Category median

Sovereignty: 1 of 6 dimensions proven

0–5 in half steps. 5 means the rubric's top anchor is met on the evidence.

by STAR AG · www.star-group.net

Report an error on this page Is this your product? →

Link to this page

Transit: AI-assisted analysis on whats-best.ai Transit: AI-assisted analysis on whats-best.ai (dark)

Add ?theme=dark to the image address for dark backgrounds. The badge shows the category and the month of the latest evaluation, and EU-Made where that is the product’s provenance. It never shows a score.

Transit at a glance

Vendor
STAR AG
Headquarters
Incorporated in Switzerland source, captured 17 Nov 2026
Data hosting
Not stated on the vendor’s pages as of 17 Nov 2026
Free plan
Not stated on the vendor’s pages as of 17 Nov 2026
Panel verdict
The AI panel rates Transit 2.5/5, the mean of 6 judges. Transit is STAR AG's translation and localization suite, spanning Transit, TermStar, CLM WebEdit and STAR machine translation. how scores are made
As of
Vendor pages captured 17 Nov 2026; evaluated by the panel 19 Nov 2026. all sources

Transit alternatives, European ones first →

Read this page as one judge. Each weighs the same scores by what they care about.

The Skeptic

Weighted verdict

Has seen "human quality" on an engine nobody has evaluated in public, "we never store your data" next to a TMS that forwards every segment to three LLM providers, and a translation memory built over six years that could only leave as a spreadsheet. Reads that a product is "enterprise-grade secure" and asks how long the text is kept, which models it reaches, and whether the memory and termbase export as TMX and TBX — and what leaving would cost.

Same scores as the panel view — this lens weights them the way this judge cares.

Scored by The Skeptic

Panel median The Skeptic This judge’s scores over the panel median. Each label opens its reasoning below.
Scores per criterion, in stars out of 5
Criterion Panel median The Skeptic
Machine translation & AI quality 3 of 5 2.5 of 5
Translation memory, terminology & review workflow 3 of 5 3 of 5
Integrations, file formats & developer tools 2 of 5 2 of 5
Confidentiality of translated content 2 of 5 2 of 5
Data export, portability & self-hosting 2 of 5 1.5 of 5
European sovereignty 1.5 of 5 1.5 of 5

Machine translation & AI quality

How this is scored

What the machine translation does and how well it is controlled: the languages and language pairs covered, the engine (own neural or LLM model, or a choice of third-party engines such as DeepL, Google, Microsoft or an LLM), glossaries and formality applied during translation, document translation that keeps the layout (Word, PowerPoint, PDF, InDesign), custom or adaptive models trained on the customer's own translations, automatic quality estimation or scoring, and published evaluations of translation quality.

0 — No machine translation of any kind, and no way to connect an engine.

3 — Machine translation is offered or one engine can be connected; languages, glossary use and quality controls are basic or not described.

5 — Machine translation for the major European and world languages with the list published, glossaries applied during translation, a formality or tone setting, document translation that keeps formatting for common office formats, and either an own engine or a choice of several named engines.

8 — Custom or adaptive models that learn from the customer's translation memory or corrections, automatic quality estimation per segment with a threshold that routes weak segments to human review, LLM-based translation with style guides and context, engine selection per language pair stated, and a published quality evaluation with its method.

10 — Translation quality that an organisation can measure and steer: evaluations against named benchmarks or blind human review published with their method, quality estimation calibrated and explained, custom models with their training data and isolation documented, and every machine-translated segment traceable to the engine, model version and glossary that produced it.

Report an error

The Skeptic

The vendor documents machine translation both as its own offering — with a dedicated system trained on company texts, translations and terminology — and as controlled gateways to named third-party engines (DeepL, Google, Microsoft Translator), plus a web tool for gisting whole documents and post-editing through Transit's QA functions. But the captured pages give no language list, no glossary or tone application during machine translation, and no published quality evaluation, so I cannot place it higher. 1 2 3

Report an error

Translation memory, terminology & review workflow

How this is scored

How translation is managed as an ongoing process: translation memory with fuzzy matches, a termbase or glossary managed centrally, style guides, review and approval stages, automated QA checks (placeholders, tags, length, terminology, punctuation), an in-context or visual editor, screenshots and context for translators, tasks and deadlines, roles per language, and ordering or inviting human translators and agencies inside the product.

0 — No management of translations at all; text goes in and comes out with nothing stored or reviewed.

3 — Translations can be edited or a glossary kept; translation memory, review stages and QA checks are basic or not described.

5 — A translation memory reused across projects, a central glossary, at least one review or approval stage, automated QA checks for placeholders and terminology, comments and screenshots for context, and roles that limit translators to their languages.

8 — Configurable multi-step workflows with automation rules, an in-context or visual editor for apps or websites, QA checks configurable per project with blocking rules, style guides applied by the AI, translation memory and termbase shared across projects with penalties and priorities, vendor or agency management with ordering inside the product, and progress and cost reporting per language.

10 — A translation process an enterprise runs on: workflows per content type with conditional routing on quality scores, linguistic quality assessment with a named framework (MQM or similar), termbase with approval workflow and forbidden terms, full revision history per segment, and productivity and quality metrics per translator and vendor.

Report an error

The Skeptic

Terminology is clearly managed: a company-wide pool, disallowed-term highlighting, term suggestions on the fly, entry verification, import and synchronization of terminology databases, and morphological support for over 80 languages; the web review tool adds memory and machine-translation suggestions, reviewer approval, segment comments, page-layout PDF context, revision history and configurable rights. I found no public information on configurable multi-step workflows with automation rules, on ordering translators or agencies inside the product, or on progress and cost reporting per language. 2 4 5

Report an error

Integrations, file formats & developer tools

How this is scored

How content gets in and out: the file formats supported (XLIFF, PO, JSON, YAML, Android XML, iOS strings and xcstrings, Office documents, InDesign), a REST API and CLI, SDKs and over-the-air updates for apps, Git repository sync (GitHub, GitLab, Bitbucket) and CI pipelines, connectors for CMS, shop and design tools (WordPress, Contentful, Shopify, Figma), browser and office add-ins, webhooks, and a JavaScript snippet or proxy for website translation.

0 — No documented way to get content in or out except typing or pasting by hand.

3 — File upload and download or a basic API; formats, repository sync and connectors are basic or not described.

5 — A documented REST API, a list of supported file formats covering the common software and document formats, sync with at least one Git host or CI, and connectors or plugins for several named CMS, shop or design tools.

8 — A CLI and SDKs for named platforms, over-the-air updates for mobile or web apps, Git sync with branching, webhooks, connectors for the main CMS, shop, design and support tools, office or browser add-ins, a public status page, and API limits documented.

10 — Integration that scales across an organisation: an API covering every function of the product with versioning and SDKs, an app or connector framework third parties build on, continuous localization in CI with pull requests opened automatically, and every connector documented with its limits and the fields it syncs.

Report an error

The Skeptic

Support for all common formats (Office and layout formats, PDF, HTML, XML, localization) is claimed, an API into third-party CAT environments is named, and connectors exist for Adobe Experience Manager, Salesforce and COTI-standard transfer to and from content management systems. I found no public information on REST API documentation, a CLI, SDKs, Git repository or CI sync, webhooks, or connector limits and synced fields. 2 3 4 6

Report an error

Confidentiality of translated content

How this is scored

What happens to the text a customer translates: whether source and target texts are stored and for how long, deletion after translation, whether customer content is used to train the vendor's or a third party's models, which machine translation and LLM providers a TMS sends text to and whether that can be restricted or switched off, encryption in transit and at rest, single sign-on, roles and audit logs, and certifications (ISO 27001, SOC 2) with their scope. Where the text is processed is scored under sovereignty.

0 — Customer texts are stored indefinitely and used for the vendor's own purposes, or no statement on data handling exists at all.

3 — The product is described as secure or GDPR-compliant and a privacy policy exists; retention of translated texts, model training and third-party engines are basic or not described.

5 — A documented statement that customer texts are not used to train models, retention or deletion of translated texts stated, the third-party engines or LLM providers named, encryption in transit and at rest, single sign-on on a business tier, and a data processing agreement offered.

8 — Texts deleted immediately after translation or after a retention period the customer sets, third-party AI processing switchable per project or account, zero-retention terms with the named LLM providers, audit logs, granular roles, ISO 27001 or SOC 2 certification stated with its scope, and a private or dedicated deployment option.

10 — Confidentiality that can be verified: an on-premise or customer-controlled deployment, customer-managed encryption keys, every subprocessor that sees content listed with its role, independent audit reports available, and every data flow of a translation request documented end to end.

Report an error

The Skeptic

There is one sentence I actually like — "Your texts will only be used for your own translations" — plus single sign-on, an in-house deployment option and third-party engines named up front (DeepL, Google, Microsoft Translator). But the captured privacy statements govern the website (Matomo, newsletter, Brevo, payment providers), not the translation pipeline: I found no public information on how long translated texts are kept, encryption at rest, a data processing agreement, audit logs, or ISO 27001 / SOC 2 certification. 2 3 7 8

Report an error

Data export, portability & self-hosting

How this is scored

Whether translation assets stay the customer's and can leave with them: export of translation memory and termbase in open formats (TMX, TBX, XLIFF, CSV), export of all projects and files, import from another TMS, an API that can read every asset, the return or deletion of data on termination, open-source code and a self-hosted or on-premise edition.

0 — Translations cannot be exported; they exist only in the vendor's interface or engine.

3 — Translated files can be downloaded; export of translation memory and glossary, import and termination terms are basic or not described.

5 — Translation memory exportable as TMX and glossary as TBX or CSV, all project files downloadable, import of translation memory and glossaries from another tool, and an API that reads translations and keys.

8 — Complete export of every asset (memory, termbase, files, comments, history) in open formats, a documented migration path from named competitors, the data returned or deleted on termination under stated terms, and a self-hosted or on-premise edition with its own documentation.

10 — The customer owns the assets end to end: open-source code under a named licence that runs fully self-hosted with feature parity stated, a complete export another tool can import without loss, and the deletion of every copy on termination documented.

Report an error

The Skeptic

The terminology pages describe complex import and synchronization functions for terminology databases, and on-site and floating licences exist, which suggests assets can be consolidated rather than typed back in. I found no public information on exporting translation memory or termbases as TMX, TBX, XLIFF or CSV, on exporting all projects and files, or on what is returned or deleted at termination — a six-year memory should be able to leave as more than a spreadsheet, and nothing here says it can. 2 3 4

Report an error

European sovereignty

How this is scored

Who the contracting entity and the parent company are, where texts are translated, stored and processed, and who the subprocessors are (hosting, machine translation engines, LLM providers, CDN for over-the-air delivery, support and payment tools). Independently sourced by the sovereignty pipeline; translated texts often hold personal data and trade secrets, so the whole chain counts, including the engines a TMS calls. A self-hosted edition the customer runs on its own European servers counts as the customer's chain, whatever the vendor's seat.

0 — Non-EU vendor and contracting entity, hosting unstated, subprocessors unnamed, and no data processing agreement offered.

3 — A privacy policy and a data processing agreement exist, but hosting location is unstated or non-EU and the subprocessors are not listed.

5 — EU hosting available and an EU contracting entity or EU representative, but parts of the chain (machine translation or LLM providers, CDN, support, payment) run through non-EU providers without a stated safeguard.

8 — EU hosting on named infrastructure, EU contracting entity, subprocessors listed with their locations including the engines and LLM providers, a data processing agreement published, and security certifications (ISO 27001, SOC 2) stated — or a self-hosted edition documented that keeps every text on the customer's own servers.

10 — Sovereign end to end and evidenced: European ownership, entity, hosting, engine and every subprocessor, or a self-hosted open-source edition with no call home, data export and deletion documented, and certifications published with their scope rather than implied.

Report an error

The Skeptic

The controller is STAR AG in Ramsen, Switzerland, with a data protection officer at STAR Deutschland in Sindelfingen, and the machine translation gateways name their engines (DeepL, Google, Microsoft). But where translated texts are hosted or processed is not stated, I found no public information on a data processing agreement, and the only listed subprocessors serve the website — Brevo for newsletters, Matomo's InnoCraft in New Zealand, Google Ireland for video — while the subprocessors of the translation chain itself are not listed. 2 3 7 8

Report an error

Pricing transparency not rated — the vendor publishes no price

How this is scored

Whether a buyer can compute the real annual cost from public pages alone: price per tier by characters, words, keys, hosted words or seats, what happens when the volume exceeds the tier, the number of users, projects and languages included, which features (machine translation volume, AI credits, SSO, integrations, API access) cost extra, the free tier and its limits, the self-hosted edition and its support, minimum term and setup.

0 — No public prices at all; every customer is a sales call.

3 — A starting price or a free tier exists, but how cost scales with volume, seats or features is unclear.

5 — Tiers are public with their price per volume band and main inclusions, but at least one commonly needed piece (overage, extra seats, machine translation or AI volume, SSO, connectors) is unpriced or "on request".

8 — Every tier priced publicly with its scaling rule (characters, words, keys, seats), overage handling stated, add-ons and AI credits priced, the self-hosted edition's cost or licence stated, minimum term and VAT treatment stated.

10 — Complete price computability: annual cost derivable for a given volume, seat count and feature set, with every add-on, setup, support and overage fee and the cancellation and data-return terms published.

Report an error

The Skeptic

Every captured page routes the buyer to a "non-binding translation quote" request or an email to the vendor; I found no public prices, no tiers and no scaling rule for volume or seats. The licence models — on-site, floating, time-limited, IaaS and SaaS — are named, which is the only pricing structure a buyer can see before the sales call. 1 3 4 6

Report an error

European sovereignty — proven facts

1 of 6 dimensions proven

Built only from facts shown on the vendor's own pages. A dimension we could not prove is left open, not scored as zero.

Hosting location Not determined — uncited Report an error
Hosting options Not determined — uncited Report an error
Open source Not determined — uncited Report an error
Data processing agreement Not determined — uncited Report an error
Transfer impact assessment Not determined — uncited Report an error

Questions about Transit

Is Transit European?

Not determined: the vendor’s pages as captured do not say where Transit is based. Its vendor, STAR AG, is incorporated in Switzerland, per the vendor’s page captured 17 Nov 2026.

Where does Transit store data?

Not stated on the vendor’s pages as of 17 Nov 2026.

Does Transit have a free plan?

Not stated on the vendor’s pages as of 17 Nov 2026.

What are the best alternatives to Transit?

As of 19 Nov 2026, the panel rates LingoHub (3.5/5), Crowdin (3.5/5) and SimpleLocalize (3.5/5) highest among the other 22 Translation & Localization products evaluated. Ratings that look equal are rounded to half stars; the order uses the unrounded rating, then the name. The highest-rated European alternative is LingoHub (EU-Made, Austria). Ratings are the AI panel’s opinion, formed from public evidence: how scores are made. The full list, with European products first: Transit alternatives.

Where this could be wrong

What we left out

A claim that does not survive our checks costs us the claim, not the page. This is what was taken off this one.

Sources (8)

The pages every claim on this page was read from — each one checked, dated, and kept verifiable.

  1. 1 Vendor page www.star-group.net Checked 17 Nov 2026 Details →
  2. 2 Machine translation & AI quality — found from sitemap www.star-group.net Checked 17 Nov 2026 Details →
  3. 3 Machine translation & AI quality — found from sitemap www.star-group.net Checked 17 Nov 2026 Details →
  4. 4 Translation memory, terminology & review workflow — found from sitemap www.star-group.net Checked 17 Nov 2026 Details →
  5. 5 Translation memory, terminology & review workflow — found from sitemap www.star-group.net Checked 17 Nov 2026 Details →
  6. 6 Integrations, file formats & developer tools — found from sitemap www.star-group.net Checked 17 Nov 2026 Details →
  7. 7 Confidentiality of translated content — found from sitemap www.star-group.net Checked 17 Nov 2026 Details →
  8. 8 Confidentiality of translated content — found from sitemap www.star-group.net Checked 17 Nov 2026 Details →