Book a demo
For enterprise
The Translation Management System Enterprises Run Their AI On
Choose the AI models. Set the quality rules. Keep your translation memories, glossaries and suppliers. wxrks runs the work and keeps a record of every decision.
Named a Leader on G2for Translation Management Systems
Independently audited(SOC 2 Type 2) since 2020 Hosted in the US
or the EU
1Choose the AI models
3Keep your memories
2Set the quality rules
Global teams run translation on wxrks
“With this one simple front door, we actually removed maybe 98% of the friction that we had in the past.”
Elitza Dublewa-Servatius
Manager Digital EX & Change, Coca-Cola Europacific Partners
SlatorCon 2026
41,000employees
30+markets
Model Management
Choose the AI Model Behind Every Task
Pick the provider and model for each AI task, for the whole account or for one business unit. Use the models wxrks runs for you, or bring your own keys.
- OpenAI, Anthropic, Google, Groq, Cerebras and more, changed in one setting.
- Models keep nothing after each request, and nothing trains on your content.
- Tokens are spent where you decide. Exact matches from your memory and repeated sentences never reach a model.
- Token use by task and by business unit, with a forecast for the month.
Your knowledge stays yours
Nothing to Retrain, So Nothing to Lose
What your team knows lives in your account: memories, glossaries, instructions and every reviewer decision. The model reads them when it translates and keeps nothing. Change providers tomorrow and every decision your reviewers made still applies.
“We got many offers suggesting to pay for a model training per language pair. And for me, this wasn't sustainable.”
Elitza Dublewa-Servatius, Coca-Cola Europacific Partners
Quality as a gate
Quality Is a Risk to Gate, Not a Score to Report
An AI-first program doesn’t ask whether a translation is good. It asks which sentences carry risk, which rule applies to each, and whether the rule was applied.
Example
A 4,000-word product page
English to Thai
Smells checks every sentence
against your glossary, memory and instructions
Score above your threshold
Nothing risky found
A simple fix
Tags, punctuation, spacing
Needs judgment
Meaning, terminology, tone
What your reviewer reads
12 sentences, each with the issue named
Not 4,000 words read against the source.
What the manager sees
2 terminology issues fixed, 1 tone note accepted with the reason
No need to read Thai to know the page is ready.
Your rules
Set the Rules Once, Checked on Every Sentence
Smells, the wxrks quality check, reads each translation against your glossary, memory and instructions, and names what it finds. You set the rules per account, unit or project.
- Set the confidence score a sentence needs to be confirmed automatically.
- Choose how strict Smells is: low, standard or high.
- Save your brand voice once in a Sous Chef Profile and apply it by account, unit, project or language pair.
Connected and visible
Your Systems In, Clear Progress Out
Content comes in from the tools you already use. Your stakeholders see where every language stands, and finance sees what it costs.
19 integrations
Contentful, WordPress, Webflow, Drupal, Adobe Experience Manager, Zendesk, GitHub, GitLab, Bitbucket, Figma, Marketo, Eloqua and more. New and changed content arrives on its own.
Progress your CMO can see
Where every language stands, for each project and the whole portfolio, without asking the team.
Quotes and payables in one place
Quotes are ready when work arrives, supplier payables are tracked, and cost goes to the business unit that asked for the work.
Built for large teams
Every Brand, Region and Supplier in One Account
Organizational units keep brands, regions and suppliers apart under one roof. Each unit has its own memories, glossaries, workflows, AI models, prices and members.
Unlimited users and units
Thousands of people in one account, each seeing only their own work.
Set once per unit
Instructions and reference files are copied into every new project in the unit.
Your suppliers inside
Agencies and translators work in your account, with no per-seat fees.
Cost by unit
Progress per language and cost charged to the unit that raised the work.
Security
Ready for Your Security Review
Independently audited
SOC 2 Type 2, held since 2020.
Your region
AWS in the United States or the European Union (Ireland), your choice.
Your sign-in
Single sign-on with Okta, OneLogin and Microsoft Entra ID.
Your data stays yours
No customer content trains any model. It is written into the data processing agreement.
New in 2026
Built This Year for Teams Running Many Languages
In October we shipped our first redesign since 2022: clearer status on every project and a quieter editor. These are the 2026 releases enterprise teams use most.
Continuous Projects
One project stays open, and only changed content becomes new work.
Work Packages
Send a whole project to another wxrks account, memory and glossary included.
Your AI assistants, connected
Use your wxrks memories and glossaries from Claude, Cursor and other AI tools.
Auto Task Placement
Tasks go to eligible linguists on their own, checked against rates, languages and workflows.
Figma preview
See each translation on the Figma design before it goes back.
Quality Reports in bulk
Export many reports at once, tracked changes included.
Switching to wxrks
Bring Your Memories, Glossaries and Suppliers
Import
Your translation memories and glossaries come in at any size, tuned until matches behave the way they do today.
Pilot
Every language in scope runs end to end with your own reviewers before you switch.
Go live
A named customer success manager, a weekly status report and quarterly reviews. The first live project usually runs within two to four weeks.
From our customers
In Our Customers’ Words
“Overall the new UI update is amazing, featuring a great, calm technology interface, and a ton of useful features.”
Sasan Banavawxrks user, reviewed us on G2LinkedIn profile
“Companies just run their content through LLMs and think the job’s done. No workflows, no audit trail, no TMs or glossaries, no consistency. And frighteningly frequently: no expert human oversight.”
JaneCraft, part of Omnicom“It allows us to move from a defensive stance of fixing mistakes to a progressive stance of amplifying creativity.”
Rik GrantGlobal Head of Language & Cultural Services, Tag (a dentsu company)Read Tag’s announcement
Questions
Questions From Enterprise Evaluations
Can we use our own AI models?
Yes. Choose the provider and model for each task and each business unit, with wxrks defaults or your own keys.
Is our content used to train AI models?
No. Models keep nothing after each request, and no customer content trains any model, by wxrks or by any provider. It is in the data processing agreement.
Do we have to replace our translation vendors?
No. wxrks sells the platform, not translation. Your agencies and translators work inside your account.
Where is our data hosted?
On AWS in the United States or in the European Union (Ireland), your choice. A dedicated single-tenant environment is available.
How is wxrks priced?
By the volume you process, in one annual subscription. Your team, requesters and suppliers are not charged per seat. In enterprise plans, implementation, onboarding, training and asset import are included, so there is no first-year spike and no separate implementation invoice.
Book a demo
See It on Your Own Content
Start with a demo. Then run a 30-day proof of concept on your real content, at no cost.
- Your target languages and up to three of your current suppliers, working as they would in production.
- Four measures agreed at the start, including how much your reviewers edit.
What happens next
- 1Pick a time for a 15-minute call on the next screen.
- 2A quick, informal look at your use case. No preparation needed.
- 3We set you up on your own content.
