Dictate Everything
nommo vs. Willow Voice
Dictate everything, with latency you can trust. Willow markets sub-200ms but delivers four to five seconds through the cloud. nommo delivers about 100ms on short phrases from a fully local pipeline.
macOS 14 Sonoma or later. Apple Silicon. No account required.
Feature-by-feature.
Every row is a concrete capability, not a marketing claim. Yes means the feature is shipping and verified.
| Feature | nommo | Willow Voice |
|---|---|---|
| LLM cleanup | ✓ | Proprietary |
| App-aware formatting | ✓ | Limited |
| 100% local processing | ✓ | — |
| Custom vocabulary | ✓ | Team only |
| One-time purchase | ✓ | — |
| No screenshot capture | ✓ | ✓ |
| Works offline | ✓ | — |
| Sub-500ms latency | ✓ | Disputed |
| Apple Silicon optimized | ✓ | — |
What Willow Voice does well
- Private Mode on by default, stronger than most cloud tools
- Quiet Mode transcribes whispered speech, genuinely unique
- Full QWERTY iOS keyboard with voice
- Fast iteration from a small team
- YC and angel credibility from Dharmesh Shah and Alexis Ohanian
Where nommo is different
- Audio never leaves the device. Not a mode, the architecture.
- Honest latency: about 100ms measured end-to-end on short phrases, not a component
- Dual-method text injection handles apps where Willow fails
- Personal custom dictionary ships standard, not gated to teams
- Push-to-talk and toggle-record, no background recording
Dictate everything.
Free download. 14-day trial. macOS 14 Sonoma or later, Apple Silicon. No account required.