Smartcat
Provenance unknown Report an errorPanel rating · 6 judges · How to read the stars
Category median
Sovereignty: 3 of 6 dimensions proven
0–5 in half steps. 5 means the rubric's top anchor is met on the evidence.
by Smartcat, Inc. · www.smartcat.com
Report an error on this page Is this your product? →
Link to this page
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.
Smartcat at a glance
- Vendor
- Smartcat, Inc.
- Headquarters
- Incorporated in the United States source, captured 10 Nov 2026
- Data hosting
- In the EU by default source, captured 10 Nov 2026
- Free plan
- Not stated on the vendor’s pages as of 10 Nov 2026
- Panel verdict
- The AI panel rates Smartcat 3/5, the mean of 6 judges. Smartcat, a translation and localization platform metered in Smartwords (one per source word per target language), scored strongest on machine translation, where scores cluster at 7: 280+ languages, six named engines routed per language pair with a Google NMT fallback, glossary terms corrected after translation, and a per-segment quality score that flags weak segments for review. how scores are made
- As of
- Vendor pages captured 10 Nov 2026; evaluated by the panel 11 Nov 2026. all sources
Read this page as one judge. Each weighs the same scores by what they care about.
The panel's verdict
Smartcat, a translation and localization platform metered in Smartwords (one per source word per target language), scored strongest on machine translation, where scores cluster at 7: 280+ languages, six named engines routed per language pair with a Google NMT fallback, glossary terms corrected after translation, and a per-segment quality score that flags weak segments for review. It scored weakest on sovereignty and pricing transparency, both ranging 3 to 4. The contracting entity is Smartcat Platform Inc. at a private-mailbox address in Wilmington, Delaware; data is "primarily stored on our EU server" yet the privacy policy states it is duplicated to servers in China or the USA depending on the customer's region; and the four plan tiers (Adapt, Accelerate, Anticipate, Autonomous) appear without public prices, so an annual cost cannot be computed. The widest spread is in workflow and review, from 6 to 8: the in-house document translator scored 8, citing graded memory tiers, the LQA review mode and a 500,000-strong linguist marketplace, while the data protection officer scored 6, noting no public information on an in-context visual editor.
Speaks for it
- Choice of six named translation engines per language pair, with a Google NMT fallback, across 280+ languages
- A documented REST API across 30 modules, with a CLI, webhooks, 40+ connectors and 80+ file formats
- A marketplace of 500,000+ translators, editors, proofreaders and agencies with payment automation on one monthly invoice
- Translation memories export as TMX for other CAT tools, and content is returned or destroyed within 30 days of a written request
- Metering is published — one Smartword per source word per target language, with 100% memory matches and repetitions free on current plans, unlimited seats, and a 15-day trial of 15,000 Smartwords without a credit card
Held against it
- No public prices for the plan tiers Adapt, Accelerate, Anticipate or Autonomous, or per Smartword, so an annual cost cannot be computed from the captured pages
- The contracting entity is Smartcat Platform Inc. at a private-mailbox address in Wilmington, Delaware, with no company register or registration number given
- Data is primarily stored on an EU server but, per the privacy policy, duplicated to servers in China or the USA depending on the customer's region
- No public subprocessor list with locations, including where the named translation engines process texts
- Translation memory export is whole-memory only, shared memories export only by the owner, and edited bilingual DOCX files cannot be imported back
Best for
- You need 280+ language coverage with a chosen engine per language pair and vetted post-editors hireable inside the platform
- You run a continuous localization pipeline and want a documented REST API, CLI and 40+ connectors instead of manual file exchange
- Your volumes reward free 100% memory matches and repetitions plus unlimited seats at no extra cost
- You manage many clients or brands and want multiple translation memories, up to 250 per profile, attached per project
Avoid if
- You must compute a full annual cost before committing — the plan tiers are named without any public prices
- You require an EU-only contracting and storage chain — the entity is a Delaware company and the privacy policy states data is duplicated to servers in China or the USA depending on your region
- You rely on re-importing edited bilingual DOCX files or exporting part of a translation memory
- You need a dedicated formality setting — formality works only through system prompts on the LLM engines
The scores
| Criterion | Panel median |
|---|---|
| Machine translation & AI quality | 3.5 of 5 |
| Translation memory, terminology & review workflow | 3.5 of 5 |
| Integrations, file formats & developer tools | 3.5 of 5 |
| Confidentiality of translated content | 3 of 5 |
| Data export, portability & self-hosting | 3 of 5 |
| European sovereignty | 2 of 5 |
| Pricing transparency | 2 of 5 |
Machine translation & AI quality
Show reasoningHide reasoning
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.
The Localization Manager
Machine translation covers 280+ languages with a real choice of engines — OpenAI GPT, Gemini, Anthropic plus Google NMT, DeepL and Microsoft Translator — engine selection and a fallback per language pair, glossaries applied with an automatic glossary-term correction step, and an AI that adapts from our team's edits. Every AI-translated segment gets a Translation Quality Score that sends weak segments to review, which is exactly the steering I want. The accuracy figures are marketing claims without a published evaluation method, and formality is only available through a system prompt on the LLM engines, so I stop short of the top level. 4 5 1 2
The App Developer
Engine routing is genuinely steerable: each language pair gets its engine with drag-and-drop priority and a Google NMT fallback that reruns failed jobs, LLM engines (OpenAI GPT, Gemini, Anthropic) sit alongside Google NMT, DeepL and Microsoft Translator, and a profile decides which memories and glossaries feed each run, with a glossary-term QA fix step after the machine pass. The AI adapts to team edits in real time and a per-segment Translation Quality Score routes weak segments to review while auto-confirming the rest. What keeps it below the top: formality only works through the system prompt on LLM engines, and the 92%-accepted and 95%+ accuracy claims come without a published evaluation method. 2 4 5
The In-House Document Translator
Six engines are named — Google NMT, DeepL, Microsoft Translator, OpenAI GPT, Gemini and Anthropic — with the engine set per language pair, a priority order and a Google NMT fallback, glossary terms and do-not-translate entries applied during translation, scanned PDFs run through OCR, and over 80 file formats including Word, PDF and InDesign; the system also learns from my corrections and rates every segment, routing the weak ones to review. Formality is only reachable through an LLM system prompt rather than a dedicated setting, and we found no public information on a published quality evaluation with its method — the 95%+ accuracy and 92% acceptance figures come without one. 1 2 4 5
The Data Protection Officer
Machine translation covers 280+ languages through a choice of named neural and LLM engines — Google NMT, DeepL, Microsoft Translator, OpenAI GPT, Gemini, Anthropic — selectable per language pair with priority ordering and a Google NMT fallback. Glossaries and translation memories apply during translation, the system continuously learns from team edits, and a Translation Quality Score rates every machine-translated segment, auto-confirming strong ones and flagging the rest for review, with an automatic glossary-term correction step after AI translation. We found no published evaluation with its method behind the 95%+ accuracy and 92% acceptance figures, and formality instructions work only with the LLM engines via system prompt. 1 2 4 5
The E-Commerce Marketer
For my German, English and four spring-season languages this covers everything: 280+ languages, six named engines from DeepL to OpenAI GPT routed per language pair with a Google NMT fallback, and an AI that learns from my edits. A per-segment quality score routes weak segments to review, and glossary terms are corrected automatically after translation with brand and product names markable Do Not Translate — though formality for German du/Sie works only through LLM system prompts, and we found no published quality evaluation with its method behind claims like 95%+ accuracy. 1 2 4 5
The Skeptic
Engine choice per language pair with drag-and-drop priority and a Google NMT fallback, six named engines including OpenAI GPT, Gemini and Anthropic, 280+ languages, glossaries enforced with an automatic post-translation term fix, and a per-segment quality score that sends weak segments to review — that is most of the way to the top. But the quality figures ("95%+ AI accuracy", "92% of words accepted without a single human edit") are bare claims with no published method behind them, which is exactly where I stop short. 1 2 4 5
Translation memory, terminology & review workflow
Show reasoningHide reasoning
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.
The Localization Manager
This covers most of what my team runs on: translation memory with full match tiers from context matches down to low fuzzies, automatic writeback, multiple memories organized by client or brand, a central glossary applied during translation, multi-stage workflows with vendor assignment, automated QA for spelling, punctuation, terminology, consistency and formatting, and a marketplace with 500,000+ linguists plus payment automation on one monthly invoice. The dedicated LQA review mode with error-type and severity labels is a genuine linguistic assessment step. We found no public information on an in-context or visual editor, QA checks configurable per project with blocking rules, or style guides applied by the AI. 12 6 7 8 1 5
The App Developer
Memory handling is proper CAT-grade: graded match bands from 50–74% fuzzy up to 101%+ context matches, memories that build and write back automatically and stay searchable and editable, plus glossary terms and do-not-translate markers applied before the engines run. QA checks cover spelling, punctuation, terminology, consistency and formatting with a custom dictionary and permission rules for ignoring errors, there is a dedicated read-only LQA review mode with error-type and severity labels, multi-stage workflows with vendor assignment per stage, and a marketplace of 500,000+ linguists with payment automation and one monthly invoice. We found no public information on an in-context visual editor for app screens or on style guides applied by the AI. 5 6 7 8 12
The In-House Document Translator
The translation memory grades matches from context 101%+ down to low fuzzy 50–74% and builds itself as segments are confirmed, the glossary is checked and mistranslated terms corrected automatically after the machine pass, and QA flags spelling, punctuation, terminology, consistency and formatting in the editor. Multi-stage workflows assign vendors per stage with automated assignment modes, a read-only linguistic review mode labels every comment by error type and severity, and a marketplace of 500,000+ translators can be ordered and paid inside the product on one monthly invoice. We found no public information on style guides applied by the AI or a full revision history per segment. 1 5 6 7 8 12
The Data Protection Officer
A translation memory builds itself automatically, can be organized per client or brand up to 250 per profile, and offers match bands from context matches down to low fuzzies, alongside a central glossary and multi-stage workflows with vendor assignment per stage. Automated QA flags spelling, punctuation, terminology, consistency and formatting errors, a read-only LQA review mode labels errors by type and severity, and linguists can be ordered from a 500,000-strong marketplace with payment automation inside the platform. We found no public information on an in-context or visual editor, style guides applied by the AI, or QA checks configurable per project with blocking rules. 1 5 6 7 8 12
The E-Commerce Marketer
The memory builds itself as my team works with match bands from context to low fuzzy, glossaries are enforced with an automatic QA fix step, and spelling, punctuation, terminology, consistency and formatting checks flag errors right in the editor. With no translator on staff I can hire a vetted post-editor per language inside the product and pay everyone with one monthly invoice, and a read-only linguistic review mode scores errors by type and severity. We found no public information on style guides applied by the AI or QA blocking rules per project, which keeps this below the top level. 1 4 5 6 7 12
The Skeptic
Fuzzy match tiers from 50% up to context matches, translation memories that build themselves and can be searched and corrected, five automated QA checks with rules on who may ignore an error, a read-only linguistic review mode with error-type and severity labels, multi-stage workflows with vendor assignment, and a marketplace of 500,000+ linguists paid on one monthly invoice. We found no public information on a visual in-context editor, per-language cost reporting, forbidden terms, or a named quality assessment framework, which keeps it below the enterprise top levels. 1 5 6 7 8 12
Integrations, file formats & developer tools
Show reasoningHide reasoning
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.
The Localization Manager
A documented REST API with roughly 30 modules, a CLI, webhooks with custom headers and error inspection, 80+ file formats and 40+ named connectors — GitHub, Jira, Figma, Contentful, Zendesk, WordPress, Google Drive and Sheets among them — give me the continuous localization pipeline I need. The website translator covers pages without manual export. We found no public information on SDKs, over-the-air updates for apps, Git sync with branching, a public status page or documented API limits. 8 9 5 1
The App Developer
The API is the real deal for a dev team: 30 documented modules covering projects, documents, memories, glossaries, invoices, even audit logs, a CLI listed in the API documentation, and webhooks with custom headers plus an endpoint to inspect the last delivery errors. Sync exists for GitHub and Jira, there are 40+ connectors including Figma, Contentful, Zendesk, WordPress and Sitecore, and 80+ file formats with JSON and HTML escaping handled on assembly. We found no public information on SDKs for named platforms, over-the-air updates for mobile apps, the pull-request behaviour of the GitHub sync, a public status page or documented API limits. 1 5 8 9
The In-House Document Translator
A documented REST API in two versions with thirty modules, a command-line tool, webhooks with custom headers and error inspection, and forty-plus named connectors — among them GitHub, Jira, Figma, Contentful, Zendesk, WordPress, Google Drive, Docs and Sheets — get content in and out, with translated files returned in their original format plus PDF export and batch ZIP download. We found no public information on over-the-air updates for apps, a public status page, or documented API limits. 1 5 8 9 13
The Data Protection Officer
A documented REST API spans 30 modules from translation memories and glossaries to webhooks with custom headers and error inspection, plus a CLI and 80+ file formats including InDesign, PDF, XLIFF and SCORM, with scanned PDFs handled by OCR. Forty-plus connectors are named, among them GitHub, Jira, Contentful, WordPress, Sitecore, Figma, Zendesk, Google Docs and Sheets. We found no public information on SDKs, over-the-air updates, Git sync with branching, a public status page or documented API limits. 1 5 8 9
The E-Commerce Marketer
With no developer on staff, the connectors are what sell it to me: 40+ integrations including WordPress, WooCommerce, Figma, Contentful, Zendesk and Google Docs and Sheets, plus a REST API with 30 modules, a CLI and webhooks with custom headers. Formats span 80+ from InDesign and PDF to XLIFF and video files. I did note we found no public information on a Shopify connector — the shops named are WooCommerce and WordPress — nor on SDKs, over-the-air app updates or documented API limits. 1 5 8 9
The Skeptic
A documented REST API across 30 modules with a CLI, webhooks with custom headers and delivery-error inspection, 40+ named connectors including GitHub, Jira, Figma, Contentful, Zendesk, WordPress and Sitecore, and 80+ file formats queryable through the API. We found no public information on SDKs, over-the-air delivery for apps, a public status page, or documented API limits. 1 5 8 9
Confidentiality of translated content
Show reasoningHide reasoning
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.
The Localization Manager
Smartcat states machine-translated content is never used to train public AI models, encrypts in transit and at rest, offers SSO on enterprise plans with role-based access and an audit-logs API, holds SOC 2 Type II and ISO 27001, and destroys or returns content within 30 days of a written request under the data processing agreement. The third-party engines are named and LLM use can be switched off per translation profile. We found no public information on zero-retention terms with the LLM providers, the scope of the certifications, or a dedicated or private deployment option. 5 3 8 4
The App Developer
A strong paper trail on the basics: no training of public AI models on machine-translated content, encryption in transit and at rest, SOC 2 Type II and ISO 27001, single sign-on through Azure AD, Okta or ADFS, role-based access on enterprise plans, workspace isolation, audit logs fetchable through the API, and content returned or destroyed within 30 days of a written request under the Data Processing Agreement. The engines that see text are named, and LLM use can be switched off per profile or engines disabled entirely. We found no public information on customer-set retention periods, zero-retention terms with the LLM providers, or the stated scope of the certifications. 3 4 5 8
The In-House Document Translator
The promise I need in writing is there: machine-translated content is never used to train public AI models, content is destroyed or returned within 30 days of a written request under the data processing agreement, encryption covers transit and rest, SOC 2 Type II and ISO 27001 are stated, sign-in runs through Azure AD, Okta or ADFS, roles govern who views, translates, reviews and approves, linguists' download rights can be restricted for sensitive documents, and audit logs are retrievable; LLM use can also be switched off per profile. Read as a contract clause, the no-training promise names public models, and we found no public information on zero-retention terms with the LLM providers, a subprocessor list with locations, or a dedicated deployment option. 2 3 5 8
The Data Protection Officer
The captured pages state that machine-translated content is "never used to train public AI models", that content is destroyed or returned within 30 days of a written request under the Data Processing Agreement, that SOC 2 Type II and ISO 27001 apply with encryption in transit and at rest, and that single sign-on runs through Azure AD, Okta or ADFS. The engines and LLM providers are named, LLM use can be switched off per profile, audit logs are exposed via API, and enterprise plans set who can view, translate, review and approve by team, role or region. As a data protection officer I read the no-training statement narrowly — it covers only public models — and we found no public information on zero-retention terms with the LLM providers, a private or dedicated deployment, or a subprocessor list with roles. 3 4 5 8
The E-Commerce Marketer
The pages state content is never used to train public AI models, everything is encrypted in transit and at rest, SOC 2 Type II and ISO 27001 are claimed, enterprise plans add single sign-on via Azure AD, Okta or ADFS with role-based access, and content is destroyed or returned within 30 days of a written request. The third-party engines are named and LLM use can be toggled off. We found no public information on customer-set retention periods, zero-retention terms with the LLM providers, or a dedicated deployment option. 3 4 5 8
The Skeptic
SOC 2 Type II and ISO 27001 are stated, encryption in transit and at rest, SSO via Azure AD, Okta or ADFS, role-based access on enterprise plans, audit logs through the API, a self-serve data processing agreement, and a promise that machine-translated content is "never used to train public AI models" — I note the word "public", which leaves anything else unstated. Retention appears only as destruction or return within 30 days of a written request; we found no public information on a customer-set retention period, zero-retention terms with the named LLM providers, the certification scopes, or a dedicated deployment option. 3 4 5 8
Data export, portability & self-hosting
Show reasoningHide reasoning
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.
The Localization Manager
Translation memory exports as TMX or XLSX — the entire memory only, and only the owner for shared memories — projects export in TMX, XLIFF, bilingual DOCX and multilingual CSV, all files download in their original format, and memory imports from TMX, Excel or bilingual documents, which is enough to move my assets in. Content is returned or destroyed within 30 days under the agreement. We found no public information on termbase export in TBX, export of comments or history, a migration path from named competitors, or a self-hosted edition. 12 13 1 5
The App Developer
Getting out looks feasible: translation memory exports as TMX, described as working with other CAT tools, and as Excel; import takes TMX, Excel and bilingual documents; every project file downloads in its original format as a ZIP; XLIFF and bilingual DOCX are export options; the API can read memories, glossaries and documents; and the DPA commits to return or destruction within 30 days. The help pages note shared memories export only in full and only by the owner. We found no public information on termbase export in TBX format, exporting comments and history, or a self-hosted edition. 1 5 8 12 13
The In-House Document Translator
Memories export in TMX or XLSX — whole memory only, the vendor states there is no supported way to export part of one — XLIFF, bilingual DOCX and multilingual CSV come out of the editor, all project files download as a ZIP in their original format, migration in is covered through TMX, Excel and bilingual documents, and everything is returned or destroyed within 30 days on request. We found no public information on glossary export in TBX, export of comments or history, or a self-hosted edition, and the vendor states edited bilingual DOCX files cannot be imported back. 1 3 8 12 13
The Data Protection Officer
Translation memories export as TMX with compatibility settings for other tools and import from TMX, XLSX and bilingual documents; all project files download in their original formats as a ZIP batch, and API modules read memories, glossaries and document exports. Under the Data Processing Agreement content is destroyed or returned within 30 days of a written request. We found no public information on glossary or termbase export in open formats, export of comments or history, or a self-hosted edition — and memory export is whole-only, with edited bilingual DOCX files not importable back. 5 8 12 13
The E-Commerce Marketer
My translations stay portable: memories export as TMX or XLSX — though only the entire memory — files come back as a ZIP in their original format alongside XLIFF and bilingual DOCX, and everything is returned or deleted within 30 days on termination. Import covers popular translation memory formats for migrating in. We found no public information on glossary export formats, export of comments and history, a migration path from named competitors, or a self-hosted edition. 1 3 8 12 13
The Skeptic
Translation memory exports as TMX — the whole memory only — and as XLSX, translations export as XLIFF, bilingual DOCX and multilingual CSV, project files download in their original formats as a sorted ZIP, and memory imports from TMX, Excel and bilingual documents. But edited bilingual DOCX cannot be imported back, non-owners cannot export a shared memory, and we found no public information on termbase export as TBX, on export of comments and history, or on a self-hosted edition; the 30-day return-or-destroy clause is the only leaving term we saw. 5 8 12 13
European sovereignty
panel opinion
Show reasoningHide reasoning
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.
The Localization Manager
The contracting entity is Smartcat Platform Inc. at a Delaware mailbox address, data is primarily stored on an EU server but duplicated to servers in China or the USA depending on the customer's region, and hosting runs on Tier IV AWS and Microsoft Azure data centers in the US and EU. A data processing agreement with model contractual clauses is offered. We found no public information on a subprocessor list with locations — including the machine translation and LLM providers — or on an EU entity or EU representative, so the whole chain cannot be assessed as European. 2 3 9
The App Developer
The chain is only partly European: the contracting entity is Smartcat Platform Inc. at a Delaware private-mailbox address with no company register or registration number given, and while data is primarily stored on EU servers in Tier IV AWS and Azure facilities, the privacy policy states it is duplicated on servers in China or the USA depending on the customer's region. The engines that see translated text are named — OpenAI GPT, Google, Microsoft, Anthropic, DeepL — a DPA with model contractual clauses is offered, and SOC 2 and ISO 27001 are stated. We found no public information on an EU contracting entity, EU representative or a self-hosted edition that would keep texts on our own servers. 2 3 5
The In-House Document Translator
The contracting entity is Smartcat Platform Inc. of Wilmington, Delaware, at a private-mailbox address with no company register or registration number given, while my contract drafts would sit primarily on an EU server yet be duplicated onto servers in China or the USA depending on my region, in Tier IV AWS and Microsoft Azure data centres. A data processing agreement link and model contractual clauses for transfers exist, but we found no public information on an EU representative, a subprocessor list with locations, or transfer impact assessments. 2 3 9
The Data Protection Officer
The contracting entity is Smartcat Platform Inc. of 2810 N Church St PMB 50985, Wilmington, Delaware — a private-mailbox address — and we found no public information on an EU entity or EU representative. Data is "primarily stored on our EU server" on Tier IV AWS and Microsoft Azure infrastructure, but the privacy policy states it is duplicated to servers in China or the USA depending on the customer's region, and the captured pages list no subprocessors with their locations, including where the named engines process texts. A data processing agreement link and model contractual clauses exist, which is a stated safeguard, but for member letters and HR documents this is not a chain I could defend to the board. 2 3 9
The E-Commerce Marketer
As a Munich shop this is my main hesitation: the contracting entity is Smartcat Platform Inc. in Delaware, and while data is primarily stored on an EU server, the privacy policy states it is duplicated on servers in China or the USA depending on the customer's region. A data processing agreement link and model contractual clauses exist, but we found no public information listing subprocessors and their locations, including which machine translation providers see my product text. 2 3 9
The Skeptic
Primary storage is on EU servers in Tier IV AWS and Azure data centers and a data processing agreement with model contractual clauses is offered. But the contracting entity is Smartcat Platform Inc. at a private-mailbox address in Wilmington, Delaware, data is duplicated to servers in China or the USA depending on the customer's region, and the OpenAI, Google, Microsoft, DeepL and Anthropic engines are named as features, not as subprocessors with locations — no subprocessor list appears. 2 3 4
Pricing transparency
Show reasoningHide reasoning
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.
The Localization Manager
There is a 15-day trial with 15,000 Smartwords and no credit card required, the metering unit is clearly explained — one Smartword per source word per target language, with exact translation-memory matches consuming nothing — and seats are unlimited. But the four current plan names appear without any prices, and we found no public information on per-tier pricing, what happens when volume is exceeded, or which features cost extra; the only figure anywhere is the marketing claim of a cut "from $200–$300 per 1,000 words down to $1.20". I could not compute an annual cost for my volumes from the public pages. 5 12 1
The App Developer
The metering unit is explained better than most: one Smartword per source word per target language, exact matches and repetitions free on the current plans (Adapt, Accelerate, Anticipate, Autonomous), unlimited seats at no extra cost, and a 15-day trial with 15,000 Smartwords and no credit card. But we found no public information on the price of each tier or per Smartword, overage handling or add-on costs on the captured pages — the only figure is the claim "From $200–$300 per 1,000 words down to $1.20" — so a real annual cost cannot be computed from public pages alone. 1 4 5 12
The In-House Document Translator
The meter is public and intelligible — one Smartword per source word per target language, with 100% and context matches and repetitions free on the current plans Adapt, Accelerate, Anticipate and Autonomous, unlimited colleagues at no extra cost, and a 15-day trial of 15,000 Smartwords without a credit card. But we found no public price for any tier or per Smartword — only the marketing figure "From $200–$300 per 1,000 words down to $1.20" — and no public information on overage handling, add-on prices, or what single sign-on costs. 1 4 5 12
The Data Protection Officer
Costs are metered in Smartwords — one per source word per target language — with 100% memory matches and repetitions free on the current plans (Adapt, Accelerate, Anticipate, Autonomous), unlimited seats at no extra cost, and a 15-day, 15,000-Smartword trial without a credit card; there is no free plan. We found no public information on the price of each plan or per Smartword, overage handling, add-on or SSO pricing, minimum term or VAT treatment, so a real annual cost cannot be computed from the captured pages. The figure "from $200–$300 per 1,000 words down to $1.20" reads as a savings claim rather than a rate card. 1 4 5 12
The E-Commerce Marketer
The metering model is genuinely friendly to a growing catalogue: one Smartword per source word per target language, exact memory matches of 100% or higher consuming nothing on current plans, unlimited seats at no extra cost, and a 15-day trial with 15,000 Smartwords and no credit card. But we found no public prices for the named plan tiers Adapt, Accelerate, Anticipate and Autonomous, no overage or per-Smartword rate, and no minimum term. The $1.20 per 1,000 words figure sits in a savings claim rather than a price list, so I cannot compute my annual cost from these pages alone. 1 5 12
The Skeptic
The metering unit is public — one Smartword per source word per target language, with 100% and higher matches free on the current plans and unlimited seats — and there is a 15-day trial of 15,000 Smartwords without a credit card. But we found no public prices for the named plans (Adapt, Accelerate, Anticipate, Autonomous), no volume bands, overage, add-on or term terms, so a buyer cannot compute a year's cost; the "from $200–$300 per 1,000 words down to $1.20" figure is a savings claim, not a rate card. 1 4 5 12
European sovereignty — proven facts
3 of 6 dimensions provenBuilt only from facts shown on the vendor's own pages. A dimension we could not prove is left open, not scored as zero.
| Legal entity |
Incorporated in US
unverified
Why unverified?The stated address is a private-mailbox (PMB) drop in Delaware and no company register or registration number is given. |
0/2 pts | 2 Report an error |
|---|---|---|---|
| Hosting location |
EU by default
unverified
Why unverified?Data is additionally duplicated to servers in China or the USA depending on the user's or customer's region, so storage is not EU-only. |
2/4 pts | 3 Report an error |
| Hosting options | Not determined | — | uncited Report an error |
| Open source | Not determined | — | uncited Report an error |
| Data processing agreement |
Published, self-serve
unverified
Why unverified?The excerpt shows only the navigation link in the legal section, not the agreement's text or the process for concluding it. |
1/1 pts | 3 Report an error |
| Transfer impact assessment | Not determined | — | uncited Report an error |
Questions about Smartcat
Is Smartcat European?
Not determined: the vendor’s pages as captured do not say where Smartcat is based. Its vendor, Smartcat, Inc., is incorporated in the United States, per the vendor’s page captured 10 Nov 2026.
Where does Smartcat store data?
Smartcat stores customer data in the EU by default, though not exclusively, per the vendor’s page captured 10 Nov 2026.
Does Smartcat have a free plan?
Not stated on the vendor’s pages as of 10 Nov 2026.
What are the best alternatives to Smartcat?
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: Smartcat alternatives.
Where this could be wrong
- Evidence ages. The oldest capture behind this page is from 10 Nov 2026. Vendors change pricing and policies without notice; every fact reflects its source as of the capture date shown in the registry.
- Weak sourcing — Data processing agreement. The excerpt shows only the navigation link in the legal section, not the agreement's text or the process for concluding it.
- Weak sourcing — Legal entity. The stated address is a private-mailbox (PMB) drop in Delaware and no company register or registration number is given.
- Weak sourcing — Hosting location. Data is additionally duplicated to servers in China or the USA depending on the user's or customer's region, so storage is not EU-only.
- AI can misread a source. Extraction and judgement are automated; a citation guarantees traceability, not infallibility. If something here is wrong, say so — no account needed, every report is decided within 5 business days, and accepted corrections are published.
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.
- 3 compliance facts could not be confirmed on the vendor’s page as captured and were left out of this page and of the panel’s material. Know more? Tell us
- 3 product facts could not be confirmed on the vendor’s page as captured and were left out of this page and of the panel’s material. Know more? Tell us
- 2 hosting facts could not be confirmed on the vendor’s page as captured and were left out of this page and of the panel’s material. Know more? Tell us
- 2 integrations facts could not be confirmed on the vendor’s page as captured and were left out of this page and of the panel’s material. Know more? Tell us
- 1 data fact could not be confirmed on the vendor’s page as captured and was left out of this page and of the panel’s material. Know more? Tell us
- 1 legal fact could not be confirmed on the vendor’s page as captured and was left out of this page and of the panel’s material. Know more? Tell us
Sources (13)
The pages every claim on this page was read from — each one checked, dated, and kept verifiable.
- 1 Vendor page www.smartcat.com Checked 10 Nov 2026 Details →
- 2 Legal notice — found from the homepage www.smartcat.com Checked 10 Nov 2026 Details →
- 3 Privacy policy — found from the homepage www.smartcat.com Checked 10 Nov 2026 Details →
- 4 Machine translation & AI quality — found from sitemap help.smartcat.com Checked 10 Nov 2026 Details →
- 5 Machine translation & AI quality — found from sitemap www.smartcat.com Checked 10 Nov 2026 Details →
- 6 Translation memory, terminology & review workflow — found from sitemap help.smartcat.com Checked 10 Nov 2026 Details →
- 7 Translation memory, terminology & review workflow — found from sitemap help.smartcat.com Checked 10 Nov 2026 Details →
- 8 Integrations, file formats & developer tools — found from sitemap developers.smartcat.com Checked 10 Nov 2026 Details →
- 9 Integrations, file formats & developer tools — found from sitemap www.smartcat.com Checked 10 Nov 2026 Details →
- 10 Confidentiality of translated content — found from sitemap www.smartcat.com Checked 10 Nov 2026 Details →
- 11 Confidentiality of translated content — found from sitemap www.smartcat.com Checked 10 Nov 2026 Details →
- 12 Data export, portability & self-hosting — found from sitemap help.smartcat.com Checked 10 Nov 2026 Details →
- 13 Data export, portability & self-hosting — found from sitemap help.smartcat.com Checked 10 Nov 2026 Details →