Your Own AI Transcription Studio,
Right on Your Desktop.
No complex setup needed. High-accuracy speech recognition, editing, and AI analysis — all in one app. One-time purchase, no monthly fees. A fully offline Windows desktop app
Windows 10/11 (64-bit)
Sound familiar?
You don't want to send audio data to external servers
Your audio data never leaves your PC. All processing happens entirely on your machine
Setup and command-line operations feel intimidating
Just drag & drop your files and click a button. GPU optimization is handled automatically
Recording, transcription, editing, and summarization are all separate tools
From recording to editing, AI analysis, and subtitle generation — everything in one app
Ongoing costs keep piling up
Buy once, use forever. Pays for itself in just 1-2 months compared to a $20/month service
Existing tools produce too many errors
World-class recognition engine built in. Runs smoothly on any modern PC
Key Features
More than transcription. Recording, editing, AI analysis, and video subtitles — all in one app
High-Accuracy File Transcription
High-accuracy offline transcription. Choose from tiny to large-v3-turbo models. Supports 10 languages plus auto-detection, with output in 6 formats: TXT, SRT, VTT, JSON, CSV, and LRC. Batch parallel processing for multiple files
- 4 backends: CPU / CUDA / Vulkan / OpenVINO
- Drag & drop files and folders for batch processing
- Loop detection & watchdog for automatic error recovery
Real-time Transcription
Transcribe microphone input or PC system audio in real-time. Results are displayed as a segment list with full-text copy and TXT/SRT export. Ideal for meetings, interviews, and lectures
- Simultaneous recording and real-time recognition
- System audio capture for web meetings (Pro)
Speaker Diarization
Automatically identify and separate who said what in multi-speaker audio. Essential for creating meeting minutes and interview transcripts — results can be easily refined in the editor
- Automatic speaker count detection, or manual specification (2–10)
- Set speaker count individually per file
- 24-color auto-assignment for visual identification in the editor
Editor
A dedicated tool for efficiently editing and correcting transcription results. Features per-segment audio playback, a fully keyboard-driven workflow, and auto-recovery — dramatically speeding up the finishing process for meeting minutes and interview records
- Play, split, merge, and adjust timestamps — all from the keyboard alone
- Auto-recovery: editing state is restored even after a crash or unexpected close
- Multi-tab editing of 6 formats (SRT/VTT/JSON/CSV/LRC/TXT)
3-Mode Recording & Download
Three recording modes: microphone, PC system audio capture (whole system or specific app), and YouTube/URL download. Save in 6 audio formats (WAV/FLAC/MP3/AAC/OGG/OPUS) and send directly to transcription
- Per-process capture to record audio from a specific app
- YouTube with audio-only/video and quality selection
Local LLM (AI Analysis & Summarization)
Local AI chat running entirely on your machine. Load transcription files for summarization and Q&A. Customize prompt templates, save/restore conversation history, adjust context size — a full AI analysis environment
- Streaming responses with thinking process display
- Conversation export and history management
- Control LLM server start/stop/restart from GUI
Video Subtitle Generation
Add subtitles to videos from transcription results. Supports both hardcoded subtitles (burned in) and soft subtitles (as a track). Customize font, size, color, and position
- Hard sub (burn-in) / Soft sub (track embed)
- Detailed subtitle style customization
Smartphone Integration
[In Development] Connect WhisperApp for Android over Wi-Fi to record on your phone and send to PC, leveraging your desktop GPU for fast transcription and LLM analysis
- Easy connection via QR code
- WebSocket support for real-time progress updates
Model Management & ModelHub
Freely choose speech recognition and LLM models. Install recommended models with one click, or search HuggingFace for fine-tuned models. GPU/VRAM info is auto-detected so you can check hardware compatibility before downloading
- Recommended models for beginners; advanced users can add any model
- Auto-detects GPU/VRAM and displays hardware requirements per model
- Quantization variants (Q4/Q5/Q8/F16) to balance size and quality
Automatic Engine Updates
Check and install updates for transcription, LLM, audio processing, and other engines from within the app. Automatically selects the right build for your GPU environment
- Auto-selects builds matching your GPU setup
- One-click update with startup auto-check
Smart Backend Optimization
Automatically selects the optimal GPU backend for your hardware. Detects power source in real-time, balancing performance and power efficiency automatically
- Auto-detects NVIDIA GPU (CUDA), Intel GPU/NPU (OpenVINO), and Vulkan-compatible GPUs. Works out of the box with zero configuration
- On AC power: GPU-first priority. On battery: NPU power-saving priority — switches automatically
- Automatic fallback to another backend on GPU errors — always stays stable
- Manual backend selection also available for advanced users
- Supports 4 profiles: Performance / Balanced / Power Saving / Auto
Who is it for?
Writers & Journalists
Accurately transcribe interviews and field recordings. Perfect for drafting articles and meeting notes
Researchers & Educators
Efficiently archive audio from lectures, conferences, and fieldwork
Video Creators
Auto-generate subtitles for YouTube and podcasts. Multi-language support for global reach
Businesses & Enterprises
Create meeting minutes without sending confidential data externally. Speaker diarization identifies who said what
Try it free for 7 days
Get full access to all Pro features for 7 days, completely free. No credit card required.
| Duration | 7 days |
| Available plan | Pro equivalent (all features) |
| Credit card | Not required |
Included in the trial
- High-accuracy transcription (all models & languages)
- Speaker diarization & real-time transcription
- Local LLM chat & summarization
- Video subtitles & YouTube download
- Smart backend optimization (GPU / NPU / CPU)
A license purchase is required after the trial period. Internet connection is required at app startup during the trial.