User Experience
One hotkey. One effortless flow. No UI overhead.
OiPer Desktop
Privacy-first transcription
Press. Speak. Release. OiPer writes it everywhere.
OiPer Desktop is built for speed. A single global hotkey records the moment, then injects text into your active app with zero lag.
Live SessionLocal
Hold
Global hotkey capture
Speak
Noise-tuned local model
Release
Instant text injection
“Draft the response, keep the tone friendly, and send.”
Three pillars, engineered.
Fast workflow, native performance, and privacy by default.
Performance
Native code with GPU acceleration and low latency.
Privacy
Audio and activity logs stay on-device by default.
Local processing
Transcription stays on your machine. Nothing leaves without explicit consent.
Optional online services
Plug in your own API key to enable online cleanup, or keep it local forever.
Control every lever.
Speech and model management
Choose model size, download instantly, switch with one click.
Backend preferences
Auto detect, CPU-only, or GPU acceleration modes.
Online provider configuration
Custom base URL, API key storage, model names.
Advanced options
LLM transcription for technical or specialized language.
Benchmark reality.
30 second English sample, measured end-to-end.
OiPer Desktop1.5s
Lemonfox API3.27s
Python Faster-Whisper3.55s
OpenAI Whisper 1 API6.46s