Email and messaging
Replies, updates, follow-ups, Teams, WhatsApp, and ClickUp: dictate where the message belongs.
EYRO
Speak faster. Keep sensitive work private.
EYRO turns speech into text across your Windows workflow, with normal dictation processed on your device.
Works offline, no account required for core use, hardware acceleration on supported PCs
Local dictation
Ctrl + Win → speak → insert
Microsoft Teams
Ready
Ctrl + WinMessage
Type a message
Cursor in Teams
Hold Ctrl + Win
Speak naturally
Transcribing locally
Text inserted
Place the cursor where you want text to appear.
Across your apps
One interaction across the apps you already use
Place cursor → hold Ctrl + Win → speak → text appears. Same pattern everywhere.
Active field in Microsoft Teams
“Following up on yesterday’s discussion: here’s the revised recommendation and next steps for the client review.”
Inserted locally, ready to edit
- 01
Place the cursor
Put focus wherever you already write: email, Teams, documents, WhatsApp Web, ClickUp, or a browser editor.
- 02
Hold Ctrl + Win and speak
An on-screen status shows that EYRO is listening. Hands-free mode is available as a secondary path.
- 03
Release to transcribe and insert
Whisper processes speech on your Windows PC, then inserts the transcript into the active text field.
Productivity proof
Your voice can outrun your keyboard.
A large typing study reported 52 WPM on average. A common speaking reference is about 150 WPM. That is nearly 2.9× the raw input rate before editing and review.
Average typing
52
WPM keyboard reference
Speaking reference
≈150
WPM spoken reference
Illustrative 1,000-word raw input: ~19.2 min typing vs ~6.7 min speaking, about 12.6 minutes before correction and review.
Sources: Aalto University typing study (52 WPM average); peer-reviewed speaking-rate reference (~150 WPM). Raw input rate is not finished-work time.
Open the personal WPM calculatorPersonal calculator
Estimate raw input time, not finished documents
Adjust with your own typing and speaking pace. This ignores transcription latency, corrections, and formatting.
- Typing time
- 19.2 min
- Speaking time
- 7.1 min
- Raw input difference
- 12.1 min, 2.69×
Private & fast
Stay on-device, and spend less time waiting after you speak
Normal dictation never needs a cloud round trip. On supported hardware, GPU acceleration shortens local processing time.
Privacy path
Microphone → local buffer → local Whisper → active field. Optional online features stay clearly labeled and separately controlled.
- 01
Microphone
On device
- 02
Local buffer
Temporary
- 03
Local Whisper
Local model
- 04
Active field
Inserted here
Performance
Internal Whisper Medium test on i9 + RTX 5070 Ti: 2.27 s GPU vs 6.84 s non-GPU, about 3.02× faster processing in that run.
Non-GPU reference6.84 s
GPU accelerated2.27 s
Results are from an internal founder benchmark. Performance varies by hardware, model, audio length, backend, power mode, and warm/cold state.
Calculated effect
3.02×
faster processing in this internal test
Whisper Medium, same text, i9 + RTX 5070 Ti, two runs per path
See methodology and hardware guidanceProfessional workflows
Dictate into the work, not a sidecar app
Reports and proposals
Speak a structured first draft, then edit in the destination app instead of rebuilding context elsewhere.
Notes and research
Capture reflections, outlines, and research notes at speaking speed without opening a separate web tool.
Task updates
Turn status, issue descriptions, and specifications into text without leaving your workflow.
Personal terminology
Teach EYRO client names, product terms, and specialist vocabulary with a user-controlled dictionary.
Product control
Choose the model, the shortcut, and the boundary
Selectable local models
Choose speed and accuracy tradeoffs for your hardware without changing the privacy default.
Hotkeys that fit your hands
Hold-to-talk and hands-free shortcuts stay consistent across the apps you already use.
Local history and deletion
See where history lives, copy or delete entries, and keep control over stored transcripts.
Hardware-aware acceleration
Supported GPUs can shorten the wait after speaking. CPU fallback remains available.
Pricing
One private productivity licence. One clear price
$19.99 USD, tax included, for one Windows PC. Paddle provides the secure checkout and tax handling.
EYRO Single-PC Licence
$19.99
One-time payment, tax included. One licence key activates one Windows PC.
A perpetual licence for the purchased major version, including minor performance, reliability, and feature updates for that version.
- Full offline core
- Local history & dictionary
- CPU fallback included
- GPU acceleration on supported hardware
- Email support from CAWOSH
Several PCs or a business
Contact us
Each Windows PC requires its own $19.99 licence.
Tell us how many PCs you need. CAWOSH can help with licence delivery and deployment questions while keeping every key locked to one PC.
- One licence per PC
- Perpetual access to the purchased major version
- Minor updates included
- Email-based customer support
FAQ
Clear answers before you install
For normal offline dictation, audio is processed by the local engine. Temporary recording files are deleted after transcription attempts. Any exception (updates, model downloads, licensing, or optional online services) is explained directly on the Privacy page.
Write at speaking speed without giving up control.
Download EYRO for Windows and test it in one workflow you already do, not in a staged demo.