User Experience
Press and release. No waiting, no context switching.
OiPer Desktop
Voice to text at the speed of thought.
Hold a global hotkey to record, release to transcribe, and watch your words appear inside the app you are already using. OiPer keeps audio on-device and respects your workflow.
Local by defaultGPU accelerationOptional online polish
Live Inject
Hold
Speak
Release
“Send the update by 3 PM, keep the file name the same.”
Built on three pillars.
Experience, performance, and privacy, balanced for real work.
Performance
Native code, GPU support, and low memory overhead.
Privacy
Everything stays on-device unless you opt into services.
Quick recording and transcription
- Global hotkey press and release flow
- Instant text injection into active apps
- Local or online cleanup options
Privacy-first architecture
- Local transcription with on-device logs
- No data leaves your machine without consent
- Full control over what is stored
Flexibility and control
- Backend selection: auto, CPU, or GPU
- Multiple LLMs for accuracy boosts
- Local or online text optimization
Settings that feel deliberate.
Manage models, configure providers, and tune transcription for your vocabulary without breaking your flow.
Speech and model management
Choose sizes, download quickly, and switch instantly.
Backend preferences
Auto detection, CPU-only modes, and GPU acceleration.
Online provider configuration
Base URL, API keys, model names, and cleanup policies.
Advanced options
LLM transcription for technical or specialized content.
Speed you can feel.
Benchmarked on a 30 second English sample.
SolutionTime
OiPer Desktop1.5s
Lemonfox API3.27s
Python Faster-Whisper3.55s
OpenAI Whisper 1 API6.46s