

Aqua Voice vs Willow Voice vs Ottex AI
An honest, side-by-side comparison of features, pricing, privacy, and technology. Find out which dictation tool fits your workflow.
Quick Overview

Ottex AI
Pricing
Free (BYOK, Local models), or pay-as-you-go (Ottex Provider)
Privacy
Excellent - Choose between Local models and BYOK. You control where data goes.
Key Strength
Free for personal use with BYOK/local models (vs $8-15/mo competitors), Gemini 3 Flash (99
Aqua Voice
Pricing
$8/mo, $96/yr
Privacy
Cloud-based processing with Privacy Mode option. When Privacy Mode disabled, transcript data may be stored for product improvement. When enabled, data not stored. TLS-encrypted connections. Team plans support org-wide Privacy Mode.
Key Strength
Industry-leading speed (450ms), 99

Willow Voice
Pricing
$15/mo, $144/yr
Privacy
SOC 2 compliant, HIPAA compliant, end-to-end encryption. Zero data retention - voice data never stored on servers. Audio and transcripts stored locally on-device only. Privacy Mode available. Optional anonymized data collection for improvement.
Key Strength
Context-aware AI that reads open windows for accuracy, automatic style matching per app, 100+ language support with mid-sentence switching, whisper mode, iOS keyboard app
Pricing
| Feature | Ottex AI | Aqua Voice | Willow Voice |
|---|---|---|---|
| Lifetime Price | Free (Local models, BYOK) | N/A | N/A |
| Subscription | Free (BYOK, Local models), or pay-as-you-go (Ottex Provider) | $8/mo, $96/yr | $15/mo, $144/yr |
| Education Discount | Same pricing for all | 70% off annual plans with .edu email | ~$100/yr (student/nonprofit discount) |
Platform
| Feature | Ottex AI | Aqua Voice | Willow Voice |
|---|---|---|---|
| Platforms | MacOS, iOS | MacOS, Windows | MacOS, Windows, iOS |
| Open Source |
Technology
| Feature | Ottex AI | Aqua Voice | Willow Voice |
|---|---|---|---|
| Transcription Engines | Cloud: Whisper, Gemini 3 Flash, Voxtral. Local: Whisper, Parakeet v3 | Avalon (proprietary), uses 6 models including OpenAI infrastructure | Proprietary AI (not specified) |
| Processing Options | Cloud (BYOK), Local | Cloud | Cloud |
| Model Selection | |||
| Apple Silicon Optimized | |||
| Memory Management | Excellent (~50MB) |
Dictation
| Feature | Ottex AI | Aqua Voice | Willow Voice |
|---|---|---|---|
| Dictation Type | Realtime, Post-Dictation | Realtime, Post-Dictation | Realtime |
| Speed | <500msLocal <500ms
Cloud 500ms–2s with Gemini 3 Flash | <500msInstant Mode: 200ms startup, 450ms paste. Streaming Mode: 850ms response. | <500msSub-200ms latency claimed, sub-1 second with AI post-processing |
| System-Wide Insertion | Works in any text field via floating text box interface, then paste | ||
| Trigger Options | Hotkey, Shortcut | Hotkey | Hotkey |
| Language Auto Detection | 100+ languages with Gemini 3 Flash. Mid-sentence language switching. Best multilingual quality available. | Supports 49 languages | 50-100+ languages with automatic on-the-fly detection and mid-sentence switching |
| Context Awareness | Screenshot-based context analysis. Work modes per app (email, notes, code). Custom instructions per context. | Uses system accessibility APIs to identify relevant words in active applications. Formats output based on app context (legal brief, medical note, casual email, etc.) | Reads open window context to get technical terms, names, project-specific acronyms right. Adapts tone by app category (work, messaging, email). |
| Dictation History | Stores local history of transcripts and audio | Stored locally on-device only. Does not sync across devices. Can delete all with one click. | |
| Search |
Text Processing
| Feature | Ottex AI | Aqua Voice | Willow Voice |
|---|---|---|---|
| Voice Corrections 'Scratch that', 'actually' | |||
| Precision Typing 'New line', 'new paragraph' | new line, new paragraph | ||
| Glossaries | Up to 800 custom words/phrases. No pronunciation tuning required. | Custom Dictionary for product names, acronyms, people names, technical terms | |
| Filler Word Removal | |||
| Auto-correction | Claims 99.1% accuracy, fewer errors than Siri or Google Voice. Automatically handles punctuation and paragraph breaks. | Automatic grammar, punctuation, and structure correction as you speak | |
| Translation | via AI commands | Multilingual transcription (49 languages) but no translation between languages |
AI & Automation
| Feature | Ottex AI | Aqua Voice | Willow Voice |
|---|---|---|---|
| AI Integration | BYOK - OpenRouter, Google AI, OpenAI, Anthropic, Mistral, OpenAI competible providers | Built-in (proprietary) | Built-in (no BYOK) |
| Custom Prompts | Custom Instructions feature - specify formatting preferences for numbers, currency, percentages, dashes, contractions, sentence structure | Has preset AI Rewrite options (tighten, expand, change tone, fix grammar) but not custom prompts | |
| Inline AI Commands Unique Ottex feature | Natural language editing like 'Actually, change for example to for instance'. Can say 'Switch to Streaming Mode'. | Voice commands: 'dash', 'new line', 'bullet point' for instant formatting | |
| AI Shortcuts | |||
| Meeting Summaries |
Features
| Feature | Ottex AI | Aqua Voice | Willow Voice |
|---|---|---|---|
| Menu Bar Access | |||
| Batch Transcription | |||
| Audio File Transcription | |||
| Video Transcription | With diarization (speaker separation) | ||
| Meeting Recording | With diarization (speaker separation) |
Accessibility
| Feature | Ottex AI | Aqua Voice | Willow Voice |
|---|---|---|---|
| Apple Shortcuts | |||
| Launcher Support Raycast, Alfred | Raycast, Alfred | No native Raycast/Alfred integrations found. Can be triggered via custom workflows. | No Raycast or Alfred extensions found |
Privacy
| Feature | Ottex AI | Aqua Voice | Willow Voice |
|---|---|---|---|
| Privacy Policy | Excellent - Choose between Local models and BYOK. You control where data goes. | Cloud-based processing with Privacy Mode option. When Privacy Mode disabled, transcript data may be stored for product improvement. When enabled, data not stored. TLS-encrypted connections. Team plans support org-wide Privacy Mode. | SOC 2 compliant, HIPAA compliant, end-to-end encryption. Zero data retention - voice data never stored on servers. Audio and transcripts stored locally on-device only. Privacy Mode available. Optional anonymized data collection for improvement. |
Why Ottex AI is the smart choice
While Aqua Voice and Willow Voice have their strengths, Ottex AI gives you the best of both worlds: enterprise-grade accuracy via Gemini 3 Flash, complete privacy with Local models and BYOK, and zero subscription lock-in.
Best Value
Free forever with local models or your own API keys. Optional $9/mo premium is cheaper than competitors.
Multilingual King
Gemini 3 Flash supports 100+ languages with 99.1% accuracy and mid-sentence switching.
Inline AI
Don't just dictate. Command. "Fix grammar", "Make professional", "Translate" — all while you speak.
Free for personal use • No credit card required