Features
Everything you need to replace your keyboard with your voice.
Works in Every App
Yyap runs as a system-level service, intercepting your hotkey no matter what application is in focus. Slack, VS Code, Google Docs, Terminal — if there's a cursor, Yyap can type into it.
- System-level hotkey that works across all apps
- Automatic focus detection — knows where to insert text
- Compatible with web apps, native apps, and electron apps
- No browser extensions or plugins needed
AI-Powered Cleanup
Raw speech is messy. Yyap uses AI to remove filler words (um, uh, like), fix grammar, add punctuation, and format your text so it reads like polished writing.
- Automatic removal of filler words and false starts
- Grammar correction and punctuation insertion
- Smart capitalization and formatting
- Preserves your natural voice and word choice
S|
Voice Commands That Actually Work
Say "delete the last sentence" and it disappears. Say "make that a bullet list" and it reformats. Yyap understands natural editing commands mixed into your dictation.
- "Delete that" / "undo" / "start over"
- "New paragraph" / "new line" / "bullet point"
- "Replace X with Y" for inline corrections
- Commands work mid-sentence, naturally
P|
App-Aware Tone Matching
Yyap detects which app you're in and adjusts the tone of your text. Casual and concise in Slack. Professional and polished in email. Technical and precise in code comments.
- Automatic app detection via window title
- Tone profiles for messaging, email, docs, and code
- Customizable per-app rules
- Learns your preferences over time
Session Memory
Yyap remembers what you said in the current session. "Change the second sentence" or "undo the last thing" works because Yyap keeps context of your recent dictations.
- Full session context for editing commands
- Multi-turn conversations with your text
- Undo/redo across multiple dictation turns
- Context clears when you close the session
V|
Lightning Fast
Under 700ms from the moment you stop speaking to text appearing on screen. Yyap uses a streaming architecture with edge inference to minimize latency.
- Sub-700ms end-to-end latency
- Streaming transcription — text appears as you speak
- Edge-optimized inference pipeline
- Works offline for basic transcription (Pro)
<700ms
end-to-end latency