#4242nd of 121 in VoiceFor image, video and design, video and animation, voice, speech and music

whisperjav

Local subtitle generator for Japanese video with scene, VAD and ASR pipelines

Releases and problems feed (RSS)

Stars
2.5k
License
MIT
Last commit
Nov 2026
Last release
Nov 2026
Language
Python
Stars, last 29 weeks+996
30 Apr 2026: 1,492 stars7 May 2026: 1,523 stars14 May 2026: 1,560 stars21 May 2026: 1,578 stars28 May 2026: 1,600 stars4 Jun 2026: 1,634 stars11 Jun 2026: 1,663 stars18 Jun 2026: 1,689 stars25 Jun 2026: 1,710 stars2 Jul 2026: 1,731 stars9 Jul 2026: 1,811 stars16 Jul 2026: 1,984 stars23 Jul 2026: 2,023 stars30 Jul 2026: 2,042 stars6 Aug 2026: 2,064 stars13 Aug 2026: 2,094 stars20 Aug 2026: 2,131 stars27 Aug 2026: 2,158 stars3 Sept 2026: 2,185 stars10 Sept 2026: 2,213 stars17 Sept 2026: 2,236 stars24 Sept 2026: 2,257 stars1 Oct 2026: 2,285 stars8 Oct 2026: 2,306 stars15 Oct 2026: 2,332 stars22 Oct 2026: 2,366 stars29 Oct 2026: 2,392 stars5 Nov 2026: 2,429 stars12 Nov 2026: 2,462 stars19 Nov 2026: 2,488 stars

1,492 on 30 Apr 20262,488 on 19 Nov 2026

From GitHub's daily star counts.

Checked on GitHub.

Overview

WhisperJAV turns Japanese adult video audio into SRT or WebVTT subtitles on your own machine. It splits audio into scenes, applies VAD, runs an ASR engine (Faster-Whisper, OpenAI Whisper, Qwen3-ASR, anime-whisper, HF transformers, or experimental CrispASR GGUF models), then applies Japanese-specific cleanup. It also has a two-pass ensemble with merge strategies, a GUI and CLI, and optional translation via Ollama or API providers.

Who it is for: People subtitling Japanese video locally who want tunable ASR pipelines

Strengths

  • Two-pass ensemble with seven merge strategies and savable presets
  • Japanese post-processing: hallucination removal, sound-only line removal, timing repair
  • Scene detection, enhancement, VAD and ASR model are swappable per pass
  • Built-in SRT translation via local Ollama or paid APIs, resumable

Weaknesses

  • Tuned for one domain; results vary with source audio quality
  • GGUF CrispASR pipelines are experimental and Japanese only
  • Balanced mode rejects --speech-segmenter and other VAD options
  • Qwen3-ASR and anime-whisper give no native timestamps; timing comes from the VAD

What it needs

  • GPU optional
  • Needs FFmpeg, Ollama (optional, for translation)
  • Models: Faster-Whisper, OpenAI Whisper, Qwen3-ASR, anime-whisper, HuggingFace Whisper/Kotoba

What changed

From the release notes. No flag means the note does not say so, not that the upgrade is safe.

Known problems

Open now, reported in the last 90 days

Titles are the reporters' own words, taken from the project's GitHub issues and ranked by reactions. We have not tried to reproduce any of them.

Also in Voice

See all 121
Also in Voice
RankProjectScore
1voicestudioDesktop app for voice cloning, dubbing, dictation and audiobooks115.7k stars+35,283 (+44%) in 30 daysAGPL-3.098 out of 100
2handyOffline push-to-talk dictation app that pastes transcripts into any text field39.5k stars+5,163 (+15%) in 30 daysMIT93 out of 100
3whisper.cppC/C++ speech-to-text engine running OpenAI Whisper models locally55.4k stars+895 (+1.6%) in 30 daysMIT91 out of 100
#4meetilyDesktop app that records, transcribes and summarizes meetings locally Live demo ↗ (opens in a new tab)32.9k stars+999 (+3.1%) in 30 daysMIT88 out of 100
#5agentsPython framework for realtime voice and multimodal agents on LiveKit WebRTC Live demo ↗ (opens in a new tab)15.3k stars+421 (+2.8%) in 30 daysApache-2.086 out of 100