Echo — voice dictation
directly on your Mac
Built for people who spend a lot of time with text, code, and AI agents. If you keep switching tabs, catching yourself procrastinating, or losing the thought while typing, Echo removes extra steps and helps you stay focused.
The flow is simple: place the cursor, press the hotkey, dictate, and the finished improved text appears exactly where you started.
Privacy-first by design: a native macOS app with no logins, passwords, or account lock-in. Data is not routed through third-party services, logs stay local, and the only external integration is a direct connection to OpenAI through your own API key.
$19
One-time app purchase
BYOK
Bring your own key
~$1/mo
Typical OpenAI cost
Pay-as-you-go
Billed directly by AI provider
Four workflows inside Echo
Echo is organized around four clear product scenarios, so it is immediately obvious what you can do in the app besides quick dictation.
Core workflow
Dictation
Press the hotkey, speak, and get cleaned-up text back into your current workflow. This is still the fastest path for messages, prompts, notes, and quick drafting.
Core workflow
Meeting Transcription
Record meetings with microphone and system audio, keep the full conversation, and save both the recording and transcript for longer calls, interviews, and discussions.
Core workflow
File Transcription
Drop audio or video files into Echo, queue multiple jobs, and save clean `.txt` output to disk. Useful for voice notes, podcasts, calls, and recorded clips.
Next direction
Meeting Copilot
A separate workspace for live call assistance: during a meeting it can surface hints from your knowledge base for sales, technical teams, and any conversation that depends on company data.
Why Echo is API-first today
Echo stays API-first for now to avoid loading the device with heavy local models.
Older Macs are not ideal for heavy speech models
Older Intel Macs are usually not a great fit for Whisper, Parakeet, or similar local models. Even early Apple Silicon and some M2 machines can hit high RAM use, extra heat, and noticeable lag during real-time transcription.
API mode keeps the laptop faster, cooler, and simpler
With API-based transcription, Echo does not run a large model on your Mac. That means less pressure on CPU, GPU, memory, and battery, with a more predictable experience across Intel and Apple Silicon devices.
Offline is a real advantage, but not the best default for everyone
Offline dictation is attractive for privacy and independence from the internet. But for many users, especially on older hardware, local models still create more friction than value, so Echo starts with the mode that feels more reliable in daily work.