A head-to-head comparison of Aqua Voice and WhisperTyping. We make WhisperTyping, so this isn't neutral. Every claim about Aqua links to a source you can check yourself, usually Aqua's own docs. Where they have the edge, we'll say so.
At a Glance
| Category | Aqua Voice | WhisperTyping |
|---|---|---|
| Starting price (unlimited) | $8 per month, billed annually | $5 per month, billed annually |
| Free plan | 1,000 free words per account (no monthly reset) | Generous free monthly minutes |
| Logging policy | Transcripts may be stored unless Privacy Mode is enabled | Always zero logging. Transcripts and audio never stored |
| Screen context / OCR | Accessibility APIs, processed on device, sent to cloud. Off by default | Runs locally on your computer |
| AI rewriting | On by default, refines in real time | Optional. Verbatim by default, AI modes when requested |
| Platforms | macOS, Windows, iOS | Windows (native) |
Pricing
WhisperTyping is 38% cheaper. Yearly cost of unlimited dictation, billed annually.
Aqua's free Starter gives you 1,000 free words per account (no monthly reset). WhisperTyping's free tier resets monthly. Professional ($15/mo) adds AI modes, Screen OCR, templates, and phone-as-microphone. See the pricing page.
Speed
WhisperTyping's median latency is 380 ms from end-of-speech to text on screen. This is the real round-trip across all users worldwide, including users on slow or unreliable connections, not a best-case benchmark.
Aqua Voice's llms.txt publishes Instant Mode at around 450 ms and Streaming Mode at around 850 ms. Their measurement methodology isn't public, so a side-by-side chart wouldn't be apples-to-apples. If latency matters to you, try both on your own machine and time them. The free tiers make that easy.
Privacy
Privacy Mode is opt-in on Aqua
Aqua Voice's privacy policy: "For users with Privacy Mode disabled, we may securely store transcript data on our servers to the extent necessary to improve the product...Otherwise, transcript data is not collected." So Privacy Mode is the setting that turns storage off, and the policy describes what happens when it's not enabled. WhisperTyping doesn't have a Privacy Mode setting because there's nothing to enable: transcripts and audio are never stored.
Screen context
Aqua's screen-context feature ("Deep Context") is opt-in (off by default per their user guide). Credit where it's due: their YC Launch page explains it uses "system accessibility APIs (not screenshots, like some others) and is heavily processed on device before inference to preserve privacy." So no screenshots are sent. The filtered context is then sent to Aqua's cloud alongside the audio for inference.
WhisperTyping's equivalent, Screen OCR, runs entirely on your own computer. Nothing about your screen reaches our servers.
No training on user content
Credit where it's due: Aqua's Terms of Service state, in capitals: "WILL NOT USE CONTENT TO TRAIN, OR ALLOW ANY THIRD PARTY TO TRAIN, ANY AI MODELS, UNLESS YOU'VE EXPLICITLY AGREED" and "We do not allow third parties (like OpenAI or Anthropic) to use your data to train their AI models." So even when transcripts are stored, they're not used for training. WhisperTyping just doesn't store anything in the first place.
HIPAA
WhisperTyping runs on HIPAA-compliant infrastructure. Aqua Voice's public materials (homepage, privacy policy, FAQ, terms) don't mention HIPAA or a Business Associate Agreement. Their Enterprise tier offers zero data retention as a feature. If you're dictating patient information, this matters.
How WhisperTyping Handles Your Data
Zero-Logging Servers
Transcription servers never log transcripts or audio. Audio is processed in memory and discarded. Nothing is written to disk.
Screen OCR Runs Locally
The contents of your screen never leave your machine. Screen OCR processing happens entirely on your own computer.
Your Vocabulary Stays Yours
Custom vocabulary, text replacements, and templates are stored locally. They are not used to train models or shared with third parties.
Full details on the privacy page.
AI Rewriting
Aqua is built around automatic AI cleanup. Their homepage: "Speak naturally, and let Aqua's AI refine your words as you talk" and "Aqua processes every word in real time, refining phrasing, fixing grammar." For many people that's the appeal. Others prefer dictation that gives them back what they said, not a rewrite.
WhisperTyping defaults to verbatim transcription with smart punctuation. AI rewriting is available through AI Modes (Write, Rewrite, Reply), only when you ask for it. What you said is what you get, unless you opt in to something else.
Platform
Aqua Voice supports Mac, Windows, and iOS. WhisperTyping is a native Windows app: light on memory, fast to activate, and tested in hundreds of Windows applications including Word, Outlook, Chrome, Slack, browser-based EMRs, Cursor, Claude Code, VS Code, and Notion. It also works in remote desktop, Citrix, virtual machines, and with Windows accessibility integrations.
On the Professional tier, phone-as-wireless-microphone lets you use your phone as the dictation mic on computers that don't have one. Useful in remote desktop, Citrix, VDI, and locked-down enterprise machines.
When Aqua Is the Better Pick
- You're on macOS, or want one tool across Mac, Windows, and iOS
- You want AI rewriting on by default
- You use the iOS keyboard a lot for mobile dictation
- You like Aqua's Edit Mode for inline voice commands like "make this title case" or "remove filler words"
- Cloud-stored transcripts and opt-in privacy aren't a concern
When WhisperTyping Is the Better Pick
- You're on Windows, where it's native rather than ported
- You want zero-logging by default, not as an opt-in
- You want Screen OCR running on your own computer
- You want verbatim by default, with AI rewriting available on demand
- You want unlimited dictation at $5 a month
- You work in remote desktop, Citrix, VDI, or low-spec environments
- You dictate patient information and need HIPAA-compliant infrastructure
Frequently Asked Questions
Is this comparison neutral?
No. We make WhisperTyping. We've tried to be defensible by linking every claim about Aqua to a source you can verify, usually Aqua's own docs. If anything's no longer accurate, let us know and we'll update it.
Can I try both?
Yes. Both have free tiers. Aqua's Starter gives you 1,000 free words per account (their FAQ: "all accounts get 1,000 words", with no monthly reset). WhisperTyping has a permanent free tier with monthly minutes, no credit card.
Is Aqua Voice a scam?
No. It's a legitimate, well-funded YC-backed company with a real product. The points on this page are about specific product and policy choices, not the company's legitimacy.
How do I migrate my vocabulary?
Easiest way: take a screenshot of your Aqua dictionary, drop it into ChatGPT, ask for a comma-separated list. Paste that into WhisperTyping's custom vocabulary. Takes about a minute.
What if I'm on a Mac?
WhisperTyping is Windows-only. If you're on Mac, Aqua Voice is a reasonable pick. Most dictation users are on Windows, which is where we focus.
Try WhisperTyping
Install in under five minutes. Start on the free plan, or jump straight to $5 a month for unlimited transcription.