Help

Your personal space is just for you and is created automatically when you sign up. An organization is a shared space for a team — its projects, templates, prompts, glossaries and memories are shared, and every member has a role. You can belong to several organizations and switch between them from the menu at the top of the app.

Create an organization under Organizations → Create organization, then open Members → Invite user and choose a role. The invitation is valid for 7 days. If the email doesn't arrive, copy the link with the Link button next to the pending invitation and send it yourself.

Members leave with the Leave button on the Organizations or Members page. The owner can't leave — they first transfer ownership with Make owner next to another member's name in Members (and stay on as an admin), or archive the organization.

Under Profile → Delete account. This deletes your personal space and your data. If you own an organization with other members, first transfer it, remove the members or archive it.

Machine translation with Google Translate and DeepL, and language models — OpenAI GPT, Anthropic Claude and Google Gemini. Language models translate following your instructions (the prompt) and respect the glossary and style you set.

DOCX, SDLXLIFF (from Trados), TXT and CSV. Translated DOCX and SDLXLIFF files download in the same format, and each file's segments are also available as CSV or TXT.

A prompt is the instruction to the language model — for example style, terminology and format. It uses variables such as {{source_lang}} and {{glossary}}, which are filled in at translation time. A template bundles an engine, prompt, glossary, memory and steps (for example translation followed by editing) so you can apply them to a project in one click.

Yes. An organization admin adds them under Connectivity → Translation engine API keys. Translations in the organization then go through your own account with that provider.

Add terms by hand or upload a CSV / Excel file with source, target and notes columns. A glossary works in both directions — an English → Bulgarian glossary is also used when translating from Bulgarian into English.

Create a memory and import a TMX file or a Trados memory (.sdltm). During translation each sentence is looked up in the memory: on an exact match the translation is taken straight from there, with no engine. On a close match (at least 75% of the words the same) the stored translation is given to the AI engine as a starting point and it changes only the differences — those segments are marked with the match percentage.