Free & Open Source

Cut the silence.
Keep every word.

Silenci removes silence from your videos and generates perfectly synced subtitles — all locally on your Mac. Export directly to Final Cut Pro.

Download DMG View on GitHub

macOS 14+ Apple Silicon 100% Local

Other tools chop your words in half.

Most silence-removal tools split audio by time — which cuts words mid-syllable. Silenci uses a 2-pass approach to split only at word boundaries.

Other Tools

"The doctor sai|" "d he was working"

"through bein|" "g someone mobile"

Time-based split — words get chopped

Silenci

"The doctor said" | "he was working"

"through being" | "someone mobile"

Word-boundary split — clean cuts every time

1. Drop Video

Drag any video file into the app or use File Open

2. AI Analyzes

Silero VAD detects silence. Qwen3-ASR transcribes with word timestamps.

3. Edit Words

Toggle individual words on/off. Split or merge clips freely.

4. Export

FCPXML for Final Cut Pro, SRT, iTT — with synced subtitles.

Everything runs on your Mac.

No cloud. No subscription. No API keys. No data leaves your machine.

2-Pass ASR — Never Cuts Mid-Word

Pass 1 transcribes with Qwen3-ASR + ForcedAligner to get word-level timestamps. Pass 2 splits segments only at word end_time boundaries. The result: clean cuts with perfectly synced subtitles.

Pass 1: VAD → 30s chunks → ASR + Aligner → word timestamps
Pass 2: Split at word.end_time → never mid-word

Apple Silicon Native

MLX 8-bit quantized models run on Neural Engine & GPU. Near real-time inference without external GPU.

M1 / M2 / M3 / M4

Multi-Language

ASR supports Korean, English, Japanese, Chinese and 10+ more languages. App UI available in 4 languages.

🇰🇷🇺🇸🇯🇵🇨🇳

Export Anywhere

FCPXML with inline subtitles for Final Cut Pro. SRT for YouTube/VLC. iTT for iTunes. All word-boundary split.

.fcpxml .srt .itt

Zero Config Install

Download DMG, open the app. Homebrew, Python, ffmpeg, and AI models are installed automatically on first launch.

See it in action.

Analysis Settings

Configure language, model, VAD sensitivity, subtitle parameters

Real-time Progress

Track analysis and model download with byte-level precision

Final Cut Pro Import

FCPXML timeline + iTT subtitles load directly into FCP

Built with

SwiftUI Python MLX Qwen3-ASR ForcedAligner Silero VAD PyTorch JSON-RPC 2.0 ffmpeg FCPXML

Cut the silence. Keep every word.