Engine
{settings.engine === "whisper" ? "OpenAI Whisper — 99+ languages, reliable" : "Nvidia Parakeet — faster on CPU, English-focused, auto-punctuated"}
Format Mode
{settings.formatMode === "Raw" ? "Exact Whisper output, no formatting" : settings.formatMode === "Clean" ? "Grouped into paragraphs, punctuation tidied" : "Structured with lists, headings, and sections"}
Whisper Model
{modelDescriptions[settings.modelSize]}
Parakeet Model
Parakeet CTC 0.6B (int8) — ~613MB, near-instant transcription
Compute Device
GPU acceleration requires building with CUDA feature
Language
Local LLM for smart task extraction, transcript cleanup, and formatting. Runs 100% offline.
Coming soon
AI-powered cleanup and smart extraction are being rebuilt with a faster engine. Task extraction currently uses rule-based matching, which runs automatically after each recording.
Custom vocabulary to improve transcription accuracy
{#if showProfiles}Structured formats for dictation sessions
{#if showTemplates}Output Folder
{settings.outputFolder || "Default (app data)"}
Toggle recording from anywhere. Click to change.
Theme
Font Size {settings.fontSize}px
{item}
Kon v1.0 · Powered by whisper.cpp · Built by CORBEL Ltd