Best dictation software for Windows, ranked by user sentiment
37 registry tools match this list. 7 have enough public posts to rank, 2,735 posts in all, collected between Jun 19, 2026 and Sep 16, 2026.
Updated · Every registry tool whose platform list includes Windows.
The ranking
Ordered by Index score. Where two scores are closer than their margins allow, the row says so: the order among tied tools is not settled.
-
#1
Aqua VoiceHighest score
Cloud AI dictation with low latency and custom dictionary, for Mac, Windows and iOS
CloudmacOSWindowsiOSFreemiumReal-timeCustom vocabContext-aware7.1 / 10n = 606 -
#2
HandyTop on privacyTop on platform supportTop on usability
Open-source offline dictation for Mac, Windows and Linux, MIT licensed
On-devicemacOSWindowsLinuxFreeOfflineOpen sourceSystem-wide7.0 / 10n = 176 -
#3
VoiceOS
Context-aware dictation with an agent mode that acts on apps, for Mac and Windows
CloudmacOSWindowsFreemiumContext-awareVoice cmdsSystem-wide6.9 / 10n = 80 -
#4
OpenWhisprOpen source
Open-source Wispr Flow alternative with local Whisper, MIT licensed
HybridmacOSWindowsLinuxiOSFreeOfflineOpen sourceSystem-wide6.9 / 10n = 270 -
#5
Willow Voice
Context-aware AI dictation with custom vocabulary, optional HIPAA compliance
HybridmacOSWindowsiOSAndroidFreemiumContext-awareCustom vocabSystem-wide6.8 / 10n = 166 -
#6
TypelessMost discussed
AI dictation with filler removal and tone adaptation, cross-platform
CloudmacOSWindowsiOSAndroidFreemiumReal-timeContext-awareFormatting5.6 / 10n = 769 -
#7
Wispr Flow
Cloud AI dictation that types into any app, for Mac, Windows, iOS and Android
CloudmacOSWindowsiOSAndroidFreemiumReal-timeContext-awareSystem-wide5.6 / 10n = 668
Scores and facts in one table
The same ranking as plain data. "Posts cover" is the period each tool's posts actually span.
| Rank | Tool | Score | Margin | Posts | Posts cover | Positive | Negative | Platforms | Processing | Pricing | Free |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Aqua Voice | 7.1 | ± 0.2 | 606 | Aug 5 to Sep 16, 2026 | 55% | 14% | macOS, Windows, iOS | Cloud | From $8/mo | Free tier |
| 2 | Handy | 7.0 | ± 0.4 | 176 | Jun 20 to Sep 16, 2026 | 48% | 8% | macOS, Windows, Linux | On-device | Free / open-source | Free |
| 3 | VoiceOS | 6.9 | ± 0.5 | 80 | Jun 19 to Sep 14, 2026 | 45% | 5% | macOS, Windows | Cloud | From $11.99/mo | Free tier |
| 4 | OpenWhispr | 6.9 | ± 0.3 | 270 | Jun 19 to Sep 16, 2026 | 42% | 4% | macOS, Windows, Linux, iOS | Hybrid | Free / open-source | Free |
| 5 | Willow Voice | 6.8 | ± 0.4 | 166 | Jun 22 to Sep 16, 2026 | 48% | 11% | macOS, Windows, iOS, Android | Hybrid | From $12/mo | Free tier |
| 6 | Typeless | 5.6 | ± 0.2 | 769 | Aug 22 to Sep 16, 2026 | 38% | 25% | macOS, Windows, iOS, Android | Cloud | From $12/mo | Free tier |
| 7 | Wispr Flow | 5.6 | ± 0.2 | 668 | Sep 11 to Sep 16, 2026 | 33% | 21% | macOS, Windows, iOS, Android | Cloud | From $12/mo | Free tier |
What the ranked tools have in common, and where they differ
- 1 of the 7 ranked tools process speech on the device, 4 in the cloud and 2 offer both.
- 2 are free outright and 5 more have a free tier. 0 can be bought once instead of by subscription.
- 2 are open source.
- Sample sizes differ widely: from 80 to 769 posts per tool, and collection reaches back further for quieter tools, so check the "Posts cover" column before comparing two scores.
Which tool do users rate best on each topic?
Every post is labelled with the topics it discusses. For each topic this shows the most and least favorably discussed tool on this list, among tools with at least 15 labelled posts on that topic.
| Topic | Most favorable | Least favorable | Tools compared |
|---|---|---|---|
| Accuracy | OpenWhispr (65% positive, 26% negative, n = 23) | Handy (50% positive, 46% negative, n = 26) | 6 |
| Speed | Aqua Voice (87% positive, 10% negative, n = 67) | Wispr Flow (74% positive, 26% negative, n = 35) | 6 |
| Price | Handy (91% positive, 9% negative, n = 32) | Typeless (21% positive, 75% negative, n = 131) | 6 |
| Reliability | OpenWhispr (30% positive, 65% negative, n = 20) | Typeless (2% positive, 91% negative, n = 46) | 4 |
| Privacy | Handy (100% positive, 0% negative, n = 22) | Wispr Flow (33% positive, 62% negative, n = 21) | 4 |
| Platform support | Handy (92% positive, 0% negative, n = 24) | Aqua Voice (33% positive, 48% negative, n = 42) | 5 |
| Usability | Handy (95% positive, 5% negative, n = 73) | Wispr Flow (66% positive, 34% negative, n = 173) | 7 |
Also on this list, not ranked
30 more registry tools match this list but have too few qualifying posts to rank. They are listed so the ranking is not mistaken for the whole market.
| Tool | What it is | Score | Why it is not ranked |
|---|---|---|---|
| BigHand | Digital dictation workflow for legal and professional firms on Windows | Not scored | Collected posts are mostly about something else that shares the name |
| BOSS AI | Screen-aware dictation with a native iOS keyboard, also on Mac and Windows | Not scored | 4 qualifying posts, 50 needed |
| Braina | AI assistant and dictation in 100+ languages for Windows, with offline mode | Not scored | 1 qualifying post, 50 needed |
| Caster | Open-source voice commands for 13+ languages, built on Dragon, Kaldi or WSR | Not scored | No qualifying posts collected yet |
| Cursorless | Structural voice editing for code in VS Code, built on Talon | Not scored | No qualifying posts collected yet |
| DictaFlow | Hold-to-talk dictation with keystroke injection, works in Citrix and Remote Desktop | Not scored | 1 qualifying post, 50 needed |
| Dragon Medical One | Cloud medical dictation with 30+ specialty vocabularies for EHR systems | Not scored | 12 qualifying posts, 50 needed |
| Dragon Professional | Professional dictation and voice commands for Windows, from Nuance | Not scored | 6 qualifying posts, 50 needed |
| Dragonfly | Open-source speech recognition framework supporting Dragon, WSR and Kaldi backends | Not scored | No qualifying posts collected yet |
| HyperWhisper | Open-source dictation with local Whisper and optional cloud routing | Not scored | 1 qualifying post, 50 needed |
| INVOX Medical | Medical dictation in 500+ hospitals with 15+ specialty dictionaries, Spanish-first | Not scored | 1 qualifying post, 50 needed |
| Microsoft Word Dictate | Voice dictation inside Microsoft 365 apps on Windows, Mac and Web | Not scored | No qualifying posts collected yet |
| NovaVoice | Voice OS with dictation, AI assistant and agent mode for desktop | Not scored | 1 qualifying post, 50 needed |
| OpenTypeless | Open-source dictation with choice of speech and language model, MIT licensed | Not scored | 11 qualifying posts, 50 needed |
| Oravo AI | Dictation with accent correction and code-switching for non-native speakers | Not scored | No qualifying posts collected yet |
| Philips SpeechLive | Enterprise dictation workflow for legal, medical and business on Windows | Not scored | 1 qualifying post, 50 needed |
| Serenade | Open-source voice-to-code with natural speech commands for VS Code and IntelliJ | Not scored | Collected posts are mostly about something else that shares the name |
| SpeechPulse | Fully on-device GPU-accelerated dictation for Windows and Mac, no account needed | Not scored | No qualifying posts collected yet |
| SpeechWrite | Cloud dictation workflow for legal and professional transcription | Not scored | No qualifying posts collected yet |
| Talon | Hands-free computer control with dictation, scriptable in Python | Not scored | 3 qualifying posts, 50 needed |
| Voibe | Fast private dictation for Mac and Windows, fully on-device on Apple Silicon | Not scored | 8 qualifying posts, 50 needed |
| VoiceDash | AI dictation with command mode for editing by voice, cross-platform | Not scored | 1 qualifying post, 50 needed |
| Voicetypr | Fully on-device Whisper and Parakeet dictation for Mac and Windows, AGPL licensed | Not scored | 14 qualifying posts, 50 needed |
| Voicy | AI speech-to-text across all platforms in 50+ languages | Not scored | 4 qualifying posts, 50 needed |
| Weesper Neon Flow | Offline dictation in 50+ languages for Mac and Windows | Not scored | 6 qualifying posts, 50 needed |
| Whispering | Open-source Whisper GUI for desktop and browser dictation | Not scored | 8 qualifying posts, 50 needed |
| Whisperstream | Local Whisper dictation for Windows, one-time purchase, no subscription | Not scored | 3 qualifying posts, 50 needed |
| WhisperTyping | Whisper-based dictation for Windows and web with a medical tier | Not scored | 2 qualifying posts, 50 needed |
| Windows Voice Access | Full voice control and dictation built into Windows 11, fully on-device | Not scored | 2 qualifying posts, 50 needed |
| Windows Voice Typing | Built-in Windows 11 dictation, opened with Win+H | Not scored | 9 qualifying posts, 50 needed |
Questions about this ranking
What is the best dictation software for Windows?
By aggregated user sentiment, Aqua Voice scores highest at 7.1 ± 0.2 out of 10 from 606 posts, but Handy, VoiceOS, OpenWhispr and Willow Voice are within the margin of error, so they are statistically tied. Data through September 16, 2026.
Which of these tools are free?
Aqua Voice, Handy, VoiceOS, OpenWhispr, Willow Voice, Typeless and Wispr Flow are free or have a free tier, according to the registry. The comparison table shows which is which.
Which of these tools work offline?
Handy and OpenWhispr can process speech on the device, according to the registry. The others send audio to a cloud service.
How is this ranking calculated?
Every registry tool whose platform list includes Windows. Tools need at least 50 qualifying public posts to be ranked. Each post counts 10 if positive, 5 if neutral or mixed and 0 if negative, and the average is pulled slightly toward 5 so small samples cannot top the table. The ± figure is a 90% margin of error.
Why is a tool I use missing from the ranking?
Tools with fewer than 50 qualifying posts are listed below the ranking without a rank. A tool whose collected posts are mostly about something else that shares its name is not scored at all.
Scores come from public posts on X, labelled positive, neutral, mixed or negative, then averaged with a small prior so thin samples cannot top the table. The ± figure is a 90% margin. Each tool lists the dates its posts actually cover. Data through September 16, 2026. How scoring works