— 12 languages, one source of truth —
Free to begin · No card required
The old town unfolds at its own pace — a courtyard here, a hidden café there. Sit where the locals sit, try whatever the waiter recommends, and let the city come to you. The best discoveries here happen by accident, somewhere between your second coffee and a conversation you half-understand.
El casco antiguo se despliega a su propio ritmo: un patio aquí, un café escondido allá. Siéntate donde se sientan los locales, pide lo que recomiende el camarero y deja que la ciudad venga a ti. Los mejores descubrimientos aquí suceden por casualidad, en algún punto entre tu segundo café y una conversación que solo entiendes a medias.
Ask about this translation —
Create a team, add members by email, share the glossary and styleguide they will all enforce. Pool credits at the team level; documents can be scoped to specific people.
Build the glossary once at the team level — every translator and every release pulls from the same source of truth. New language pairs reference the same terms; override per locale only when regional usage diverges.
Sentence-level control at scale. A custom direction ("shorter," "more formal") reworks just the block. The team's shared glossary and styleguide feed in automatically — no out-of-band fixes that break consistency on the next batch.
A QA pass over the whole document or a selection. Glossary and styleguide enforced; inconsistent terminology and style deviations flagged. Run it before the release goes out.
Comment threads on any block. Reviewers leave feedback inline; the team resolves and reopens as the discussion lands. No more "the localization tracker doc."
Three short comparisons, then a feature row to settle it.
Put text in, get text out. Hope you like what you get. No glossary, no styleguide, no memory of what came before.
You bring the glossary, the styleguide, the tone. Multiple options per sentence. You decide what makes the cut.
By page fifty it is calling the same character three different names and forgot what sarcasm was.
Page 300 reads just like page 1. Glossary and styleguide enforced across the run.
Impressive at parties. Try fifty docs with consistent terminology. Good luck.
Who does this for a living. Upload, translate, QA every sentence, export — on time, on brand.
| Capability | Transept | Google T. | DeepL | ChatGPT |
|---|---|---|---|---|
| Voice & tone consistency | ||||
| Glossary enforcement | ||||
| Memory across documents | ||||
| Full document context | ||||
| Sentence-level editing | ||||
| Batch translate + QA | ||||
| Team collaboration | ||||
| Notion & Google Drive |
Pay for words you actually translate. Top up when you need more — extra credits don’t expire.
One credit = one source word translated in Standard mode. Premium mode uses a deeper model and costs 3 credits per source word. Manual edits, exports, comments, and review activity are free — credits are only spent when a block is translated by AI.
For your first projects
Standard mode, 1 credit per word.
For regular client work
For agencies & large catalogs
Prices are in EUR. VAT is added at checkout where applicable — EU consumers see local VAT, while EU businesses with a valid VAT ID pay the net price under reverse charge. Top up any paid plan with credit packs — 50,000 credits for €5, 200,000 for €18, or 500,000 for €40 — and pack credits never expire. Cancel anytime; 7-day money-back guarantee on your first subscription payment, no questions asked.
Two writers, one problem.
— Vitalii.I wrote 120,000 words of satirical romance fantasy — mages drowning in debt, a multiverse economy that makes Wall Street look ethical, and a love story tangled up in all of it.
— Mariia.I write semi-fictional stories about real people, built around how it felt. Most of the people I wrote about were foreigners. We met in English, spoke in English, lived those moments in English. So I wrote them that way too.
“But I sometimes wonder what version of them I missed by not writing them in their language.”Mariia
— Vitalii.The same flatness hit me from the other side. Every tool I tried preserved the words but killed the voice. Tone, rhythm, wit, emotional texture — nothing held. It all came back flatter than it went in.
— Both of us.So we did what we know how to do. We read the machine-translation papers. We benchmarked Claude, GPT, Gemini, and a handful of open models on the same 50-page chapter — judging accuracy, tone, and whether the satire still landed. None of them held all three; each failed in a different way. So we built the tool we wished had existed when we started.