Coming soon

About This Software

Modular, fully local AI interpreting for games and voice chat.

Whether you're playing untranslated overseas indie games/JRPGs, teaming up in cross-border multiplayer

co-op, or watching foreign game showcases and esports livestreams, TransBox provides low-latency

in-game subtitles and voice interpretation. All models run on your machine — download once, then work

fully offline.

FEATURES

• Floating Subtitle Bar: an always-on-top window over your game (the game must run in Borderless

Windowed mode), click-through, streaming bilingual captions — stays out of your sightlines and

won't interfere with controls

• Real-Time Speech Recognition, Machine Translation and Speech Synthesis in one chain:

hear it → read it → hear the translation

• Two-Way Voice Relay: WASAPI process isolation — two-way translation across Steam Voice and Discord co-op

• Glossaries: terms.json forces correct translations for in-game gear, skills and lore

• Themes: built-in themes, plus community subtitle themes from the Workshop

• History Log: full-text search, export, read-aloud and single-line copy for in-game dialogue and

quest text — never miss critical clues

PERFORMANCE

• Default engines run locally; download the models once and work fully offline.

• The overlay is optimized for low VRAM usage, and click-through lets your clicks reach the game.

• It is a separate always-on-top window: it does not inject into the game, read game memory or modify

• game files, so anti-cheat systems won't flag it.

WORKSHOP

Steam Workshop is supported: subscribe to community glossaries and subtitle themes, or package and

publish your own. Workshop items are data-only — currently glossaries and themes; packages containing

executable code are not accepted.

AI Generated Content Disclosure

The developers describe how their game uses AI Generated Content like this:

This utility app uses artificial intelligence models to perform real-time speech-to-text transcription (ASR), live subtitle translation (MT), and local text-to-speech synthesis (TTS) based on the user's voice or PC audio stream.