Many clients, one workspace.
Run every client from one account. Each keeps its own project, its own glossary and its own memory setting, and the engine writes the first draft, so your linguists spend their hours on the rows that need a person.
A tool per client, and a renewal calendar to match.
Unlimited projects on every paid plan, each with its own glossary and its own memory opt-out.
The first pass eats the margin before a linguist has looked at anything.
The engine drafts and the four checks run, so a reviewer starts from a row that already passed them.
Migrating a client in means a week when nobody is billing.
6 migration importers, Lokalise, Phrase, Crowdin, Transifex, Weblate and POEditor, plus 20 file formats read natively.
$49a month, flat
- 1,000,000 words a month
- Unlimited seats and unlimited projects
- All 41 integrations: 36 from Starter, 5 from Team
- Overage at $0.0008 a word, about $80 per 100k words
Where the hours actually go.
- 01
The client pushes, or you do
A connector, the CLI or a file upload. Twenty formats are read natively, and one CSV or XLSX upload takes up to 200,000 data rows.
intake - 02
The engine drafts
Exact memory hits skip the model and cost nothing, and only the rows actually translated are billed.
drafting - 03
The four checks run
Placeholders, plurals, glossary and length, on every row. A broken placeholder, plural or term gets one corrective retry and is refused if it still fails. A row that only overruns its length is delivered, flagged for review.
four checks - 04
Your linguist reviews
In context where the client has enabled it, with the rendered screen and the length budget in view rather than guessed at.
your hours - 05
The client gets it back
Through the connector, the pull endpoint, the CLI or the CDN bundle. There is no browser download of translations.
delivery
FAQs
What an agency asks before moving its first client across.