Sayboard uses 3 recognition engines + AI to turn your voice into clean, structured text. Hold Fn, speak, release. That's it.
Apple Speech + local Whisper + cloud Whisper run in parallel. Cross-validation catches errors that any single engine misses.
GPT-4o-mini corrects errors, adds punctuation, structures long text into numbered lists, and handles self-corrections.
Short phrases output instantly when local engines agree. Long text goes through GPT for structure. Smart routing, minimal latency.
Speak in any language, output in your preferred one. Supports Chinese, English, Japanese, Korean, and 95+ more languages.
Local Whisper model runs entirely on-device. Audio never leaves your Mac unless cloud recognition is needed.
System-wide input via Fn key. Works in any app — WeChat, Slack, VS Code, browsers, terminals, and more.
Long-press the Fn key anywhere on your Mac to start recording.
Talk naturally. See real-time preview as you speak.
Let go of Fn. Clean, structured text appears at your cursor.
| Sayboard | Typeless | Whisper Transcription | |
|---|---|---|---|
| Recognition engines | 3 (cross-validated) | 1 | 1 |
| AI post-processing | GPT-4o-mini | None | None |
| Auto bullet points | Yes | No | No |
| Self-correction | Yes | No | No |
| On-device model | Whisper large-v3-turbo | Whisper small | Cloud only |
| Languages | 99+ | ~30 | 99+ |
| Price | From $9.9/mo | $9.99/mo | Pay per minute |
Start free, upgrade when you need more.
Enter your code below to unlock Pro instantly — no payment needed.