V2sub 2.0 — All-in-One AI Video Subtitle, Translation & Dubbing Studio
Rebuilt with a modern Fluent 2.0 desktop UI. Powered by offline Faster-Whisper models, AI sentence segmentation, 12+ context-aware LLMs (DeepSeek, ChatGPT, Claude, Gemini, Local LLMs), local/cloud TTS voiceover, CUDA video hardsubbing, and an exclusive Real-Time Translation Player.
Download V2sub 2.0.2 & Whisper Offline Models
All editions share the same installer. Upgrade over an existing installation preserves your license & settings automatically.
Baidu Netdisk (Recommended)
Includes Windows x64, macOS ARM64 & offline models
Quark Cloud Drive
Includes Windows x64, macOS ARM64 & offline models
Google Drive (Global)
Includes Windows x64, macOS ARM64 & offline models
Built for High-Volume Video Creators, Translators & Educators
Every module in V2sub 2.0 is engineered to eliminate manual bottlenecks—from Whisper hallucination suppression to 30-line context LLM prompts and millisecond-aligned TTS dubbing.
Drag-and-Drop Batch Processing for Files & Folders
Rebuilt from the ground up with FluentQt modern UI, featuring instant dark/light theme and Chinese/English hot-switching. Process single videos, audio files, or entire directories of hundreds of videos automatically.
- Supports MP4, MKV, MOV, AVI, FLV, M4V videos, MP3/WAV audio, and standalone SRT/ASS/VTT subtitles
- Real-time progress tracking, stage timers, and batch completion reports
- Smart skip for processed files and automatic checkpoint resume on interruption
2x Faster with Large-Turbo Model & Human-Grade AI Sentence Segmentation
Includes 6 optimized offline Whisper models with NVIDIA CUDA acceleration and multi-threaded CPU boost. Exclusive "Special Optimization" and "Protected Mode" eliminate hallucinations, while AI segmentation delivers natural subtitle lines.
- Built-in Model Manager with one-click download and instant hot-reload in v2.0.2
- Large-Turbo model doubles recognition speed while retaining Large-v3 accuracy
- Recognizes 99+ languages including English, Japanese, Korean, Cantonese, Russian, French, German, and more
Context-Aware LLM Translation with Scene Memory & Video Summary
Beyond zero-config Microsoft, Google, and DeepL translators, V2sub 2.0 deeply integrates DeepSeek, ChatGPT, Gemini, Claude, local LLMs, and any OpenAI-compatible LLM provider.
- 30-line sliding context window with automatic <summary> & <scene> memory keeps terminology consistent
- Custom video background prompt, temperature control, and automatic summary.txt generation
- Built-in V2sub Auto-Domain and one-click API connectivity tester
Millisecond-Aligned AI Voiceover & One-Click Ready-to-Publish Video Muxing
Features both offline local TTS and cloud multi-voice engines. Automatically matches voice speed to subtitle timestamps for lip-sync precision, and burns custom-styled subtitles and audio directly into the video.
- Quick sample preview ("Trial Mux") and reusable templates for unattended batch video production
- Automatic NVIDIA CUDA hardware encoding boosts video muxing speed by 2-3x
- Supports Karaoke word highlighting, max-character line wrapping, and background box masking
Watch Raw Videos in 10 Seconds with Live Asynchronous Subtitle Generation
Skip the wait! Powered by an asynchronous hyper-threaded ASR + translation engine, drag any raw foreign video into the player and start watching within 10 seconds while bilingual subtitles stream ahead in real time.
- Freely seek forward or backward—the engine prioritizes audio around your current playback cursor
- Export generated subtitles to standard SRT files anytime during or after playback
- Rock-solid VLC native video output integration on both Windows and macOS
Full SRT / ASS / VTT / TXT Support for Professional Post-Production
Control bilingual ordering, Simplified/Traditional Chinese output, line-length limits, and convert single or batch subtitle files between SRT, ASS, VTT, and TXT in one second.
- Directly generate or translate styled ASS subtitles, WebVTT, and clean TXT transcripts
- One-click ASS tag stripping to reset complex styling to clean defaults
- Built-in system diagnostics (CUDA / VRAM / Network) and proxy configuration
Flexible Plans for Every Workflow
Start with the Free Edition forever, or unlock Pro, Max, and Ultra capabilities anytime with an instant license code.
Free Edition
Get started with local speech recognition and Microsoft translation for free
- All 99+ source and target languages supported
- Includes Small offline Whisper model
- Default Microsoft Translator (no proxy required)
- Generates monolingual & bilingual SRT subtitles
- NVIDIA GPU (CUDA) hardware acceleration
Pro Edition
For solo creators: unlocks Medium/Large models, DeepSeek & ChatGPT translation
- Everything in Free + audio-only (MP3/WAV) & standalone subtitle translation
- Unlocks Medium & Large-v3 high-precision Whisper models
- Unlocks DeepSeek, ChatGPT, Google Translate & Full-Speed CPU mode
- Batch folder processing (up to 10 files per batch)
Max Edition
Most popular for power users: Realtime Player, all LLMs, DeepL & Special Optimization
- Everything in Pro + exclusive V2sub Real-Time Translation Player
- Unlocks Custom LLMs (OpenAI-compatible), Local LLMs (Ollama/LM Studio), Claude, Gemini & DeepL
- Unlocks V2sub Auto-Domain & automatic video summary.txt generation
- Unlocks Model Special Optimization & Protected Mode against hallucinations
- Unlocks ASS / VTT / TXT output & single-file Subtitle Format Converter
Ultra Edition
Ultimate studio productivity: Large-Turbo 2x speed, AI segmentation, TTS dubbing & auto video muxing
- Everything in Max + unlimited batch video & subtitle processing
- Exclusive Large-Turbo (2x speed) model & AI Sentence Segmentation
- Exclusive Local/Cloud AI Voice Dubbing (with batch TTS) & Video Hardsubbing Studio
- Exclusive One-Click Pipeline: raw video in → dubbed & hardsubbed video out
- Exclusive Karaoke subtitles, max-line-length control & batch folder format converter
Complete Feature Matrix Across Editions
Scroll horizontally on mobile| Category | Feature | Free | Pro | Max | Ultra |
|---|---|---|---|---|---|
| Input & Basics | Video to Subtitle + Translation (99+ languages & bilingual) | ||||
| Input & Basics | NVIDIA GPU (CUDA) Hardware Acceleration | ||||
| Input & Basics | Audio-only (MP3/WAV) ASR & Standalone Subtitle Translation | ||||
| Input & Basics | Folder Batch Queue Processing | 单次 10 个 / Up to 10 | 单次 10 个 / Up to 10 | 无限制 (含批量字幕与配音) / Unlimited | |
| Whisper Models & Segmentation | Available Offline Whisper Models | 小型 (1款) | 小 / 中 / 大型 (3款) | 含英语专用模型 (5款) | 全系6款 (含大型-极速版 2x提速) |
| Whisper Models & Segmentation | Full-Speed Multi-Core CPU Mode (+30%~50% speed) | ||||
| Whisper Models & Segmentation | Special Optimization & Protected Mode (Anti-Hallucination) | ||||
| Whisper Models & Segmentation | AI Model Sentence Segmentation (Human-grade line breaks) | ||||
| AI LLM & Machine Translators | Microsoft Translator (Zero-config, no proxy needed) | ||||
| AI LLM & Machine Translators | Google Translate, ChatGPT & DeepSeek Translators | ||||
| AI LLM & Machine Translators | Custom LLM (OpenAI Compatible), Local LLMs (Ollama), Claude, Gemini, DeepL | ||||
| AI LLM & Machine Translators | V2sub Auto-Domain & Automatic Video Summary (summary.txt) | ||||
| Subtitle Formats & Tools | Generate & Translate ASS, WebVTT, and Plain Text (TXT) | ||||
| Subtitle Formats & Tools | Subtitle Format Converter Tool (SRT ⇄ ASS ⇄ VTT ⇄ TXT) | 单文件转换 / Single File | 支持目录批量转换 / Batch Folder | ||
| Subtitle Formats & Tools | Karaoke Word Highlighting & Max Characters Per Line Control | ||||
| Realtime Player & Media Pipeline | V2sub Real-Time Translation Player (Watch Raw Videos Live) | ||||
| Realtime Player & Media Pipeline | Local Offline & Cloud AI Voiceover TTS (with Batch Synthesis) | ||||
| Realtime Player & Media Pipeline | Video Hardsubbing Studio & One-Click Full Automation Pipeline |
Join Our User Community or Add WeChat Support
WeChat ID: aizimuji · Official QQ Group: 157222123. Get help with model selection, LLM API configuration, or license inquiries.
Need Quick Subtitle Extraction Without Installing Desktop Software?
Our new WebAssembly + WebGPU studio runs Whisper speech recognition and keyless Microsoft/LLM translation right inside your browser.