Features

Everything you need to understand anything you watch

Every AddSub feature: real-time transcription, live translation across 99 languages, speaker diarization with color-coded labels, and a floating Liquid Glass overlay. 100% on-device.

Real-time transcription

AddSub listens to any audio playing on your Mac and transcribes it as it happens. On-device neural speech recognition runs on the Apple Neural Engine — no cloud round-trip, no per-minute fees, and results appear in under a second.

Sub-second latency on Apple Silicon
State-of-the-art accuracy, tuned for Japanese and English
Auto-detect language or pin it manually
Works with any app that plays audio

Real-time transcription live on AddSub

Live translation in 99 languages

Watch anything in any language and read it in yours. AddSub detects the source language and translates continuously using Apple's on-device Translation framework.

99 speech languages supported
Language pairs, e.g. Japanese → English
Bilingual captions (original + translation side by side)
Apple Translation runs on the Neural Engine

Live translation in 99 languages live on AddSub

Speaker diarization

AddSub separates speakers and labels each line with a persistent color-coded badge (S1, S2, S3…). Interviews, podcasts, and meetings become readable transcripts with who-said-what preserved.

Persistent speaker identity across the session
Color-coded badges that match the transcript
Works with the real-time stream, not just offline files
Calibrated on-device clustering

Speaker diarization live on AddSub

Floating Liquid Glass overlay

A gorgeous native overlay floats above every app. It stays out of your way while watching, and expands on hover for the full transcript timeline.

Floating window above all apps, all spaces
Minimized capsule ↔ full timeline modes
Copy any segment, copy all, or export to SRT
Fully customizable: font size, opacity, text color

Floating Liquid Glass overlay live on AddSub

100% on-device privacy

Every feature — transcription, translation, diarization — runs locally. Your audio never leaves your Mac. There are no servers, no accounts required, and no cloud costs.