O.Translator

O.Translator translates PDF, Word, PowerPoint, Excel and EPUB files between 100+ languages while preserving headings, tables and images. Upload, review a partial preview and quote, then pay in credits to translate the full document. Glossaries, online editing and shared team projects round it out.

O.Translator is a browser based document translation service built around one promise: the translated file should look like the original. Rather than pasting text into a translation box and rebuilding the layout afterwards, users upload a document and receive a translated version with headings, tables and images held in place across supported formats. The format list covers PDF, Word, PowerPoint, Excel, EPUB and more, with individual files accepted up to 1GB. The service works across more than 100 languages and reports over 4 million documents translated for roughly 2 million users. The workflow is deliberately staged. After uploading a file and choosing a target language and translation mode, users see a partial preview of the translation alongside a layout check, then a quote for the full job. Only after confirming does the system translate the whole document, which can be downloaded once complete. Pricing runs on credits calculated from tokens, document analysis fees and the chosen mode, with OCR and higher tier models able to shift the cost. Scanned PDFs and images are handled through OCR, and an enhanced recognition option exists for pages that the standard pass struggles with. Beyond raw translation, the platform adds light localisation tooling. A glossary keeps brand names and specialist terms consistent, the output can be edited online, and projects and glossaries can be shared so a team reviews documents in a common context. Self-directed translators, academic researchers, legal and marketing teams working across languages, and publishers handling EPUB titles are the obvious audience. Against general purpose translators such as Google Translate, the pitch is format fidelity rather than raw speed. The trade off is that the service is credit metered and document centric, so short snippets and quick lookups are better served elsewhere. It also depends on third party AI providers for processing, and retention is limited: unpaid documents are deleted 15 days after task creation, paid ones after 30.