Built for speed and user control
Desktop dictation without cloud lock-in
OiPer transcribes locally by default and injects output into your active application. Online services are optional and fully user-configured.
1.5sOiPer benchmark
Auto / CPU / GPUbackend control
LOCAL TRANSCRIPTIONGLOBAL HOTKEY FLOWOPTIONAL ONLINE OPTIMIZATIONNATIVE PERFORMANCELOCAL TRANSCRIPTIONGLOBAL HOTKEY FLOWOPTIONAL ONLINE OPTIMIZATIONNATIVE PERFORMANCE
User Experience
Single hotkey motion keeps dictation immediate and intuitive.
Performance
Native architecture delivers low latency and tight resource usage.
Privacy
Local-first processing means your speech stays under your control.
QUICK RECORDING & TRANSCRIPTION
Local Processing
Your audio and logs stay on-device
Transcription runs entirely on your machine by default for full data ownership and predictable privacy boundaries.Online Services (Optional)
Enable only when you choose
Add API key, base URL, and model only if online optimization is needed. Disable the feature at any time.SETTINGS & CONFIGURATION
Speech & Model Management
- Choose speech models by size and capability
- Download and switch models quickly
Backend Preferences
- Auto detection for best available runtime
- Dedicated CPU-only mode
- GPU acceleration where supported
Online Provider Configuration
- Custom base URL support
- API key management
- Model name configuration
- Local or online text optimization control
Advanced Accuracy
- Route transcripts through LLMs when needed
- Useful for technical vocabulary and specialized content
- Examples include compact models like Gemini 2.5 Flash Lite
PERFORMANCE // 30 SEC ENGLISH AUDIO
OiPer Desktop1.5s
Lemonfox API3.27s
Python Faster-Whisper3.55s
OpenAI Whisper 1 API6.46s
OiPer achieves the fastest transcription while preserving privacy and user-level control over every processing path.