Desktop dictation for people who care about speed and privacy.
Hold a global hotkey to record, release to transcribe, and inject the text into your active app. Local by default, online only when you choose.
1.5s
30s audio benchmark
On-device
Audio and logs stay local
Hotkey
Hold → speak → release
Core philosophy
Experience, performance, and privacy built into every release.
User experience
Fast and intuitive, no workflow overhead.
Performance
Native code with GPU acceleration where available.
Privacy
Local transcription with optional online services.
Everything you need to control transcription.
OiPer gives you the choice between local and online text optimization, multiple backends, and configurable models.
Quick recording & transcription
Global hotkey, instant injection, local cleanup.
Privacy-first architecture
Everything stays on-device unless you opt in.
Flexibility & control
Choose auto, CPU, GPU, and multiple LLMs.
Settings and configuration
Every preference exposed without clutter.
Speech & model management
Easy downloads, model size control, quick switching.
Backend preferences
Auto detect, CPU-only, or GPU acceleration.
Online provider configuration
Custom base URL, API key, and model names.
Advanced options
LLM transcription for specialized language.
Performance benchmarks
30 second English audio, end-to-end latency.
Ready for a faster workflow?
Install OiPer Desktop and dictate at the speed of thought.