Skip to content

Translate Text

Instant LLM translation with auto-detection, language swap, streaming output, and TTS playback. Best for short snippets, chat-style use, and testing your LLM setup.

Walkthrough

  1. Click Translate Text in the sidebar.
  2. Type or paste your source text into the left pane.
  3. The Source language auto-detects as you type (powered by langdetect).
  4. Pick a Target language from the right-side dropdown.
  5. Click Translate (or press Ctrl+Enter).
  6. The translation streams into the right pane token-by-token.

What you get

  • Streaming output — translation appears as the LLM generates it, no waiting for the whole response.
  • Auto-detect source — the source picker updates in real time. Click it to override.
  • Edit mode — click the right pane to edit the translation manually. Press Escape to cancel an in-flight translation; press it again to exit edit mode.
  • History reuse — every translation is saved. Click an entry in the Text Translation History panel below to re-load both panes; edits update the original entry instead of creating a duplicate.
  • TTS playback — click the Listen button next to either pane to hear it read aloud. Honours your Settings → Voice → TTS method pick — Edge TTS (default), ElevenLabs, Google Cloud TTS, Gemini TTS, or Piper TTS (fully offline). With Piper selected, the Listen button runs the same pre-flight as the Voice page: a missing per- language voice surfaces a modal dialogue with an Open Settings button so you can download it. Cache hits skip the pre-flight entirely.
  • Per-feature model picker — when more than one LLM is configured, a dropdown lets you pick a fast Flash model for speed or a heavier Pro model for quality, just for this page.

Shortcuts

Shortcut Action
Ctrl+Enter Translate
Ctrl+L Swap source ↔ target
Escape Cancel in-flight translation, or exit edit mode
Ctrl+F Focus the history search

Tips

RTL languages

Translations into Arabic, Hebrew, or Persian automatically render right-to-left in the output pane. The same RTL handling carries through to file output across every format on the Translate Document page (PDF, DOCX, PPTX, XLSX, ODF, RTF, HTML, EPUB, ASS/SSA), and Persian gets a native fa-IR voice for Edge TTS playback.

Listen-button cache

The first time you hit Listen on a given (text, language) pair the audio is synthesised and cached on disk. Subsequent plays are instant. The cache is wiped at app startup, so each session starts fresh.

Where the keys go

The Translate Text page reads the same keychain entries as the rest of the app — see LLM Providers.