AI Translate¶

A free, cross-platform desktop translator that handles 45 languages and goes far beyond plain text — it translates documents, audio, video, live speech, screen captures, and more, all with a single LLM-powered pipeline.

Desktop app

Drag a file in, pick a target language, get a translated copy back. Drag-and-drop, history, glossaries, the works.

5-minute walkthrough
Command line

ait report.docx --target French — the same pipeline, scriptable and headless. Useful for CI, batch jobs, servers.

CLI guide
AI agents (MCP)

Expose translation as Model Context Protocol tools so Claude Desktop, Claude Code, and other MCP clients can call them directly.

MCP setup

What you can translate¶

Kind	Formats
Office docs	`.docx`, `.xlsx`, `.pptx`, `.odt`, `.ods`, `.odp`, plus legacy `.doc` / `.xls` / `.ppt`
PDFs	extract-overlay translation with layout preservation, bookmark / form / link translation, OCR fallback for scans
Text & web	`.txt`, `.md`, `.rst`, `.html` / `.htm` / `.xhtml`, `.xml`, `.rtf`, `.json`, `.csv`, `.epub`
Subtitles	`.srt`, `.vtt`, `.ass`, `.ssa`
Localization	`.po`, `.pot`, `.xliff` / `.xlf`, `.yaml` / `.yml`, `.properties`, `.strings`
Images	`.png`, `.jpg`, `.jpeg`, `.bmp`, `.webp`, `.tiff`, `.tif` (OCR or LLM vision)
Audio	`.mp3`, `.wav`, `.m4a`, `.flac`, `.ogg`, `.aac`, `.wma`
Video	`.mp4`, `.webm`, `.mkv`, `.avi`, `.mov`, `.wmv` (full dubbing pipeline)

Headline features¶

Translate Text — instant LLM translation with auto-detection, edit-in-place, TTS playback. Right-to-left languages (Arabic, Hebrew, Persian) render natively.
Translate Document — drop files, watch a per-task progress spinner, get translated copies side-by-side. RTL targets get proper bidi markup; Ctrl+P / Ctrl+G pause and continue the queue.
Generate Subtitle (STT) — transcribe audio / video into SRT / VTT / ASS / SSA.
Generate Voice (TTS) — synthesize subtitles into MP3 / WAV with timing.
Video Dubbing — full STT → translate → TTS → mix back into the source video.
Live Translation — real-time microphone or system-audio subtitle overlay.
Extract Text — OCR or LLM vision → .txt / .docx.
Glossary — enforce consistent terminology across translations.

Vertex AI mode for Gemini

Enterprise users can flip Gemini calls from the Developer API to Vertex AI in Settings → LLM — point it at your GCP project and region, optionally supply a service-account JSON path. See LLM Providers.

First time here?

Start with Installation, then the 5-minute first-translation walkthrough. You'll have a translated document in under 10 minutes from a fresh clone.