Sayboard: AI voice keyboard for iOS

Speak in any app. Sayboard turns voice into ready-to-send text — fully offline, no servers, no subscriptions. Open source. One-time purchase.

KEY FEATURES

• Voice-to-text in any app on iPhone and iPad — Messages, Mail, Slack, Notes, Safari, anywhere you type • AI tools right on the keyboard — fix grammar, change tone, rewrite, summarize • Speech-to-English translation from 90+ languages • Automatic punctuation, capitalization, and per-app writing style • Hands-free dictation while walking, driving, or commuting • Bluetooth headset support and iPad-optimized layout

WHY SAYBOARD

• 100% on-device — works in airplane mode, even in a tunnel • No accounts, no analytics, no telemetry • Source code public on GitHub — verify our privacy claims yourself • One-time purchase. No subscriptions, no ads, no in-app purchases • Choose your AI models — from ultra-light to high-accuracy

PICK YOUR ENGINE

Speech recognition — pick the model that fits your device, accuracy needs, and battery preference:

• WhisperKit (OpenAI Whisper, highest accuracy) • Parakeet (NVIDIA, fast and lean) • Moonshine (ultra-light, smallest footprint)

Text processing with local LLMs:

• Gemma (Google) • Qwen (Alibaba) • Llama (Meta)

Every model runs on your device's Neural Engine. No cloud, no API keys, no token quotas.

WHO IT'S FOR

• Writers — long-form drafts at the speed of speech • Professionals — emails, Slack, meeting notes, hands-free • Multilingual users — type in your language, send in English • Privacy-conscious users — sensitive notes that never leave the device • Developers — open source, audit-friendly, customizable

ONE-TIME, FOREVER

Pay once. Yours forever.

All future updates, models, and features included. No subscriptions. No ads. No in-app purchases. No accounts. No tracking.

OPEN SOURCE

Sayboard's full source code is on GitHub. Audit our privacy claims. Fork it. Build it from source. Trust through transparency.