Everything you need to translate
business documents at scale
Under the hood
One pipeline, every format
Every document — regardless of format — goes through the same five stages, so formatting fidelity never depends on which file type you upload.
Upload
Any of 14 formats
Extract
Format-aware parsing
Translate
Multiple AI quality tiers
Reassemble
Formatting preserved
Deliver
Same format, translated
Reassemble, proven
File formats
14 formats. One workflow.
If your team creates it, Vernacia can translate it. Each format has a dedicated parser that preserves the specific formatting properties of that file type.
Output modes
Three ways to receive your translation
Different documents need different delivery. Choose the output mode that fits your review process, not the other way around.
AI models
The right quality tier for every document
Choose based on your accuracy requirements. Tier selection is per-job — run low-priority documents on Standard and critical ones on Precision.
14 formats · 73 languages · 3 output modes
All in the same pipeline. Pick your format, language pair, and delivery mode independently — no separate tools, no re-uploading.
Language support
73 languages — and counting
Any language supported by OpenAI models. Full RTL support for Arabic, Hebrew, and Persian. Correct CJK line-breaking for Chinese, Japanese, and Korean.
+ 53 more languages including Bengali, Swahili, Tagalog, Ukrainian, Vietnamese, and more.
Teams
Built for teams that translate together
PRO and Enterprise plans include full team collaboration — shared projects, review workflows, and role-based access control.
PRO & EnterpriseFAQ
Common questions
What if the AI mistranslates a term?
Every completed job receives a 0–100 quality score, and PRO and Enterprise plans include a built-in review workflow — a teammate can approve, reject, or request revision before a translation is treated as final.
Is my document data used to train AI models?
No. Documents are sent to OpenAI's API under their Data Processing Agreement, and OpenAI does not use API-submitted data to train their models.
What's the maximum file size?
10 MB per file on Starter, 50 MB on Pro, and 200 MB on Enterprise. Contact [email protected] if you need to translate something larger.
Can I keep terminology consistent across documents?
Yes — Translation Memory automatically reuses approved translations for repeated terms, and PRO+ plans can auto-extract or import a custom glossary.
Is my document confidential?
Yes. Files are encrypted in transit (TLS 1.3) and at rest (AES-256), only accessible to your account or team, and automatically deleted 30 days after the job completes — or immediately on request.
More questions? See the full FAQ.