Voice memos,
supercharged with
AI workflows.

Talkie turns your spoken thoughts into structured, automated output — all on your devices, powered by iCloud, and designed for builders who want speed, sovereignty, and flow.

Voice Recording & Transcription

One-Tap Capture

Instant recording startup time. No lag. No loading screens.

Auto-Transcribe

Local, high-accuracy transcription running on-device via Apple's Neural Engine.

iCloud Sync

Seamless, encrypted synchronization across iPhone, iPad, and Mac.

Smart Library

Organize with Recent, Processed, Archived, and custom Smart Folders.

Widgets & Quick Access

Record From Anywhere

Add Talkie widgets to your Home Screen, Lock Screen, or Control Center for instant voice capture — no need to open the app.

Home Screen Widget

Small or medium widget with quick record button and memo count.

Lock Screen Widget

Circular widget for instant capture without unlocking.

Control Center

iOS 18+ toggle for one-tap recording from anywhere.

Deep Links

Widget taps open directly to recording view.

9:41
5G100%
Record12 memos
iOS 16+

Siri & Shortcuts Integration

Hands-Free Control

Use Siri voice commands to record memos, play recordings, search your library, and more — all without touching your device.

"Hey Siri, record with Talkie"

Instantly starts recording

"Play my last Talkie memo"

Plays most recent recording

"Search Talkie for meeting notes"

Searches transcriptions

"How many memos in Talkie?"

Quick count of your library

Shortcuts App

All Talkie actions are available in the Shortcuts app for custom automations.

  • Start Recording
  • Play Last Memo
  • Search Memos
  • Get Memo Count

Example Automation

"When I arrive at the office → Start recording with Talkie"

AI-Powered Workflows

Modular Pipelines

Build modular pipelines that process your voice memos through multiple steps. Use our drag-and-drop editor to chain LLMs, scripts, and file operations.

SUPPORTED MODELS

Gemini, OpenAI, Anthropic, Groq, Local MLX

VARIABLES

{{TRANSCRIPT}}, {{TITLE}}, {{DATE}}

Available Step Types
LLMSummaries, extraction, restructuring
Shell CommandRun CLI tools (claude, gh, jq)
Save to FileWrite results to disk with aliases
WebhookSend JSON/Text to any endpoint
EmailSend results via Mail.app
CalendarCreate events from transcript
ClipboardCopy results to system clipboard
NotificationNative macOS alerts

Shell Command Integration

Run Unix Tools Directly

  • Executable allowlist for safety
  • Claude CLI integration (MCP)
  • Multi-line templates
  • Respectful PATH merging (brew, node, bun)

# Example: Create GitHub Issue

gh issue create
--title "{{TITLE}}"
--body "{{TRANSCRIPT}}"
--label "voice-memo"

Path Aliases

Shortcuts for your most important directories.

  • @Notes~/Documents/Obsidian/Vault
  • @Projects~/Dev/Current

Smart File Output

Save workflow results exactly where you want.

  • Template filenames with date/time
  • Auto-directory creation
  • Append mode for logs & journals

Example Workflows

Voice → Obsidian

Extract insights → enrich with Claude CLI → save to Markdown in your vault.

Meeting Notes → Tasks

Extract todos → structure JSON → send to Todoist or Linear via API.

Daily Journal Builder

Summarize daily thoughts → append to daily journal file with timestamp.

Quick GitHub Issue

Dictate bug report → transform to format → gh issue create.

Security Architecture

Data Sovereignty Model v1.2
Input (iOS)
Processing (macOS)
Services
User Owned Zone

iCloud

Encrypted CloudKit Container

iPhone

Input Context
1. Audio Capture (VAD)
2. Local Encryption
3. Sync Push

Mac

Runtime Details
1. Context Assembly
2. PII Redaction
3. Secure Dispatch
External Services
OpenAIOpenAI
AnthropicAnthropic
GoogleGoogle
NotionNotion
ZapierZapier
LinearLinear

Outbound: Text-Only Stream

Only the final, sanitized text prompt is sent. Audio files never leave your device.

Platform Support

  • macOS (Primary App)
  • iOS (Companion App)
  • iCloud Sync

Configuration

  • Workflow Manager
  • Activity Log
  • Model Library
  • API Key Management