Dictate. Record meetings.
Search everything.
Everything you say — dictations, meetings, memos — captured on your Mac, organized into a daily journal, searchable by your AI agents. Nothing leaves your device.
NotionSystem-wide dictation
Talk into any app. Resonant types it.
Context-aware voice-to-text in every text field. Knows the app, window, URL, code file, and Slack recipient. 25 languages. On-device.
Works offline
No internet? No problem.
Everything runs on-device. No internet required. Airplane mode, air-gapped networks, coffee shops with bad Wi-Fi — it all works.
Learn moreNative notch UI
Feedback lives in your Mac's hardware notch.
No floating windows. No overlays. Visual and audio feedback for every state — idle, listening, processing, done, and meeting recording — all rendered in the notch.
Hi Sarah, I wanted to let you know we'll need to push the deadline back by one week. The vendor hasn't delivered the API documentation yet. I'll keep you updated. Best, Tom
Adaptive tone per app
One voice. Many tones.
Resonant adapts your dictation to how you write — formal for email, casual for Slack, technical for your IDE. Same words in, different output per app.
Composing in Gmailhey david just following up on the proposal from last week we're ready to move forward with the revised scope but need the signed SOW by friday so we can kick off the sprint on monday let me know if you need anything from our side
Email mode
Speak it raw. Send it polished.
Context-aware formatting for emails. Detects Gmail, Outlook, and mail apps — adds greetings, sign-offs, and proper line breaks automatically.
Q2 Migration Plan
The mobile team is targeting a June release. Infrastructure changes must land before feature freeze.
We need to make sure that the auth service is deployed independently from the mobile app because otherwise the mobile team will be blocked waiting for auth endpoints and that could delay the whole Q2 timeline significantly.
Alex will own the revised timeline. Target: auth endpoints live by April 15.
Selection transform
Highlight. Speak. Rewritten.
Select text, speak an instruction, text is replaced. Works with any spoken command in any app.
YouWhat if we start auth in parallel with the mobile redesign?
AlexMy concern is the mobile team needs the new endpoints first.
YouLet me talk to Priya about staggering the deploys.
Meeting recording
Records both sides. No bot. No one knows.
Auto-detects Zoom, Teams, Meet, and more. Transcribes both sides with speaker attribution. Hidden during screen shares.
Learn moreDecouple auth deploys
Today, 1:15 PM · 42s
Idea: decouple auth deploys from mobile release train. Ship auth service independently behind a feature flag. This unblocks the mobile team without coupling their timeline to ours.
Summary
• Ship auth service independently with feature flag
• Unblocks mobile team from auth timeline dependency
• Avoids coupling release trains
Voice memos
Speak a thought. It titles itself.
Capture ideas with a dedicated memo mode. Auto-title, AI summaries, and audio playback.
Auto journal
Your day, written for you.
Builds a daily work log from everything you said. Timeline, highlights, and task blocks — grouped by app.
Full-text search
Type a name. Find every conversation.
Search across every dictation, meeting, and memo by keyword. Results grouped by type with timestamps and app context — so you find the thing, not just the day.
Listening...
Corrected text appears here
Personal dictionary
Teach it your words.
Custom hot words for names, jargon, medical terms, and technical vocabulary. Compiled at startup with zero per-dictation overhead.
Listening...
Rich text output appears here
Smart Links
Say it. It links it.
Add trigger phrases that auto-replace with clickable links during dictation. Mention a Notion doc, Linear ticket, or URL — Resonant links it.
MCP for your agents
Your AI agents can query everything.
11 MCP tools. Works with any MCP-compatible agent, including Claude and Codex. Search dictations, read transcripts, pull journal entries.
Learn moreAnd everything else.
13 more capabilities built for professionals who use voice every day.
Local LLM post-processing
Route dictations through Apple Intelligence, Ollama, LM Studio, or any OpenAI-compatible endpoint with your own system prompt.
HIPAA mode
Encrypted local history, full audit log, and telemetry suppression for regulated environments. GDPR controls with one-tap deletion.
25 languages
Optimized for English with dedicated models. Fluent in 24 others including Mandarin, Japanese, Korean, Russian, Spanish, and German.
HotMic
Always-listening streaming transcription. Text appears as you speak — no button press. Pauses automatically during meetings.
Cloud cleanup
AI post-processing that fixes STT errors and adapts to context. Email mode, message mode, and general mode — zero data retention.
Long dictation
Record up to 5 minutes continuously with consistent speed. Streaming model handles extended sessions without degradation.
Live preview
Real-time transcription while you speak. Watch your words appear before you release the trigger key.
Auto-paste & auto-send
Text pastes directly into the active field. In messaging apps, optionally sends the message automatically — no manual paste or Enter.
Custom trigger key
Use fn, right Option, or any custom key to start dictation. Configure separate triggers for dictation, memo, and dispatch modes.
Progression & streaks
Track total words, dictation count, minutes saved, and streak days. Stats cards in your feed show your voice-first momentum.
Audio cues
Synthesized click, tack, and snap sounds for press, release, and paste complete. Haptic-like feedback without headphones.
Voice recording persistence
Every dictation and memo saved as a local audio file with playback in the feed. Replay what you said alongside the transcript.
Low-latency pipeline
Rust-compiled text formatter with 19 transforms at startup. Background model warm-up. Paste latency under 50ms for Electron apps.