Attachment Translation reads images, PDFs and voice messages - right-click, Translate Attachment. 5 Credits plus output, 1 Credit per started minute of speech.
Right-click (or long-press on mobile) a message with an image, PDF, or voice message attached, go to Apps, and choose Translate Attachment. Nothing posted in a channel is read automatically - it only runs when someone clicks it.
Pigona reads one attachment per request - the first supported one on the message, or the first image or PDF that fits if that one is over 7 MB - transcribes the text or speech it finds, and translates it into your personal language. The reply is a private, ephemeral message - only the person who ran the command sees it. If the message carries more than one supported attachment, the reply says which one it used, for example attachment 2 of 3.
A PDF is read up to 20 pages and about 28,000 characters of extracted text. Text in Chinese, Japanese, Korean or Thai counts for more, so those documents reach the limit sooner. Scanned or photographed pages are read as images. Longer documents, and files that are corrupted, password-protected or too complex to read in time, are refused before anything is charged, and a PDF with no readable text at all costs nothing. Pages that come back with no readable text are listed under the translation. If Pigona can only read part of a document, nothing is delivered and nothing is charged.
A voice message or Ogg Opus (.ogg) file is read up to 10 minutes long. Longer ones are refused before anything is charged. Pigona cuts out the silence on its own servers before anything is sent for translation, so only the speech is sent and billed per started minute, and a recording that is only silence costs nothing. If a recording holds sound but no speech, only the per-minute part is charged.
This page is part of the Pigona Documentation, a complete guide to setting up and using Pigona - the context-aware AI translation bot for Discord. Pigona supports 100+ languages, Autonomous Translation, Reactive Translation, Relay Channels, and Smart Context Engine.