Ottex
Mac, iOSBest when: You want one tool for system-wide dictation, meeting recording, and finished written outcomes like follow-up emails, Linear tickets, CRM updates, PRDs, and docs
Dragon is still a serious Windows dictation tool, especially for custom vocabulary, Auto-Texts, macros, and local desktop transcription. The gaps are modern meeting work, transparent pricing, macOS desktop support, and local or BYOK options across newer AI dictation tools.
Each section opens with the takeaway. The table underneath is the receipt — every cell carries its source and the date we checked it.
Dragon is the odd one out on pricing because the public pages route Professional v16, Professional Anywhere, and Dragon Anywhere Mobile through contact-sales or unpublished subscription pricing. QuickWhisper and VoiceInk are the cleanest anti-subscription picks at $59 one-time and $25-$49 one-time, while Ottex gives a broader menu with free local or BYOK paths, $0.20/hour PAYG, $14/month Pro, and $24/seat/month Team. Beware teaser free tiers: Willow Voice and Wispr Flow list 2,000 words/week, which is only about 15 minutes of dictation a week at 130 words per minute.
VoiceInk | Spokenly | |||||
|---|---|---|---|---|---|---|
| Free tier | YesSource: ottex.ai — checked Jun 2026Forever free with local models and BYOK providers. | NoNo free plan | NoNo free plan | PartialSource: spokenly.app/pricing — checked Jun 2026Unlimited use; no usage limits | PartialSource: granola.ai/pricing — checked Jun 202625 meetings/month | PartialSource: quickwhisper.app — checked Jun 2026Tiny and Base Models; free transcription of audio and video files using the Tiny and Base models, with no limitations. |
| Subscription | $14/moSource: ottex.ai — checked Jun 2026$14/mo — unlimited dictation and meeting intelligence, all hosted models included. | NoNo subscription plan | NoNo subscription plan | $9.99/moSource: spokenly.app/pricing — checked Jun 2026The pricing page shows annual billing at $99.99 per year and references Pro cloud features. | $18/moSource: granola.ai/pricing — checked Jun 2026Business is $18 per member/month or $14 per member/month billed annually. | NoNo subscription plan |
| Pay-as-you-go / one-timeEscape the flat subscription | $0.2/hourSource: ottex.ai — checked Jun 2026Hosted models at cost: dictation ~$0.2/h, meetings ~$0.5/h (~$3/mo average). | NoSubscription only | $25 one-timeSource: tryvoiceink.com/buy — checked Jun 2026Lifetime of updates | NoSubscription only | NoSubscription only | $59 one-timeSource: quickwhisper.app — checked Jun 2026Pay-once use forever. |
| Bring your own keyPay the model provider, not the middleman | YesSource: ottex.ai — checked Jun 2026OpenRouter, Google AI, OpenAI, Anthropic, Mistral, and compatible providers. | NoSource: dragon.nuance.com/en-us — checked Jun 2026No BYOK support is published. | YesSource: tryvoiceink.com/docs — checked Jun 2026Cloud and custom models may require adding the provider API key. | YesSource: spokenly.app/pricing — checked Jun 2026Users can connect their own API keys from OpenAI, Deepgram, Groq, ElevenLabs, Mistral, and other providers. | No public claim | YesSource: quickwhisper.featurebase.app/changelog — checked Jun 2026Changelog lists adding existing API keys for cloud transcription providers and AI models. |
| Education discount | NoSource: ottex.ai — checked Jun 2026Same pricing for everyone. | No public claim | PartialSource: tryvoiceink.com/faq — checked Jun 2026Student discount is available by contacting the vendor with student ID; no public percentage is listed. | YesSource: spokenly.app/pricing — checked Jun 2026Verified students get 50% off Pro after contacting support with an academic email. | No public claim | No public claim |
Dragon remains Windows-first on desktop, with iOS and Android mobile availability in the US and Canada, but no macOS desktop, Linux, web, or browser extension. Wispr Flow has the broadest dictation platform spread here with Mac, Windows, iOS, and Android, while Aqua Voice covers Mac, Windows, and iOS. Ottex, VoiceInk, MacWhisper, QuickWhisper, FluidVoice, Alter, and Hyprnote are Apple-first or Mac-only, so they are wrong answers for a Windows or Android switch reason.
VoiceInk | Spokenly | |||||
|---|---|---|---|---|---|---|
| Platforms | Mac, iOSSource: ottex.ai — checked Jun 2026 | Windows, iOS, AndroidSource: dragon.nuance.com/en-us — checked Jun 2026 | Mac, iOSSource: tryvoiceink.com/faq — checked Jun 2026 | Mac, Windows, iOSSource: spokenly.app/download — checked Jun 2026 | Mac, Windows, iOS, AndroidSource: granola.ai/download — checked Jun 2026 | MacSource: quickwhisper.app — checked Jun 2026 |
| Native app footprintLight on memory, instant to invoke | YesSource: ottex.ai — checked Jun 2026Native Swift app, ~50MB RAM. | YesSource: nuance.com/products — checked Jun 2026Dragon Professional v16 is installed as a Windows application. | YesSource: github.com/Beingpax — checked Jun 2026The GitHub README describes VoiceInk as a native macOS application. | No public claim | No public claim | YesSource: github.com/quickwhisperapp — checked Jun 2026QuickWhisper is distributed as a macOS app through direct download, App Store, GitHub, and Homebrew. |
Dragon does not record meetings, mix mic and system audio, run bot-free capture, or identify speakers in meetings. Ottex is the broad dictation-plus-meetings pick for Mac and iOS, while Vowen adds Mac and Windows bot-free capture for Zoom, Teams, Meet, Slack Huddles, and system audio. Granola, Hedy, Hyprnote, Krisp, and Otter.ai are stronger meeting specialists, but most do not replace Dragon's core dictation job.
VoiceInk | Spokenly | |||||
|---|---|---|---|---|---|---|
| Meeting recordingThe call is captured at all | YesSource: ottex.ai — checked Jun 2026With speaker separation. | NoSource: dragon.nuance.com/en-us — checked Jun 2026Not supported. | NoSource: tryvoiceink.com/docs — checked Jun 2026Not supported. | NoSource: spokenly.app/docs — checked Jun 2026Not supported. | YesSource: help.granola.ai/article — checked Jun 2026Granola records and transcribes meetings without joining as a bot. | YesSource: github.com/quickwhisperapp — checked Jun 2026Vendor lists recording meetings in Zoom, Google Meet, Microsoft Teams, Webex, and others. |
| Bot-free captureNo "recording bot" joins your call | YesSource: ottex.ai — checked Jun 2026System-audio capture, no bot joins the call. | NoSource: dragon.nuance.com/en-us — checked Jun 2026Not supported. | NoSource: tryvoiceink.com/docs — checked Jun 2026Not supported. | NoSource: spokenly.app/docs — checked Jun 2026Not supported. | YesSource: granola.ai — checked Jun 2026Granola does not send a bot to meetings. | YesSource: github.com/quickwhisperapp — checked Jun 2026Vendor says meetings are recorded without bots joining calls. |
| Speaker separation | YesSource: ottex.ai — checked Jun 2026 | NoSource: dragon.nuance.com/en-us — checked Jun 2026Not supported. | NoSource: tryvoiceink.com/docs — checked Jun 2026Not supported. | NoSource: spokenly.app/docs — checked Jun 2026Not supported for live meetings. | No public claim | YesSource: quickwhisper.featurebase.app/changelog — checked Jun 2026Vendor lists global speakers and speaker filtering across history. |
| Calendar detection | YesSource: ottex.ai — checked Jun 2026Calendar-based meeting detection. | NoSource: dragon.nuance.com/en-us — checked Jun 2026Not supported. | NoSource: tryvoiceink.com/docs — checked Jun 2026Not supported. | NoSource: spokenly.app/docs — checked Jun 2026Not supported. | YesSource: help.granola.ai/article — checked Jun 2026Granola surfaces calendar events and can start notes from upcoming meetings. | No public claim |
| Cross-meeting search | YesSource: ottex.ai — checked Jun 2026 | NoSource: dragon.nuance.com/en-us — checked Jun 2026Not supported. | NoSource: tryvoiceink.com/docs — checked Jun 2026Not supported. | NoSource: spokenly.app/docs — checked Jun 2026Not supported. | YesSource: help.granola.ai/article — checked Jun 2026Ask Granola can answer across past meetings. | YesSource: quickwhisper.featurebase.app/changelog — checked Jun 2026Changelog lists related transcriptions and semantic search across History. |
| Shared team memory | PartialSource: ottex.ai — checked Jun 2026Public share links, not a full team workspace. | NoSource: dragon.nuance.com/en-us — checked Jun 2026Not supported. | NoSource: tryvoiceink.com/docs — checked Jun 2026Not supported. | NoSource: spokenly.app/docs — checked Jun 2026Not supported. | YesSource: granola.ai/pricing — checked Jun 2026Business workspaces include shared notes and folders for teams. | No public claim |
Dragon stops well before modern meeting intelligence: no summaries, decisions, action items, transcript chat, follow-up drafts, CRM updates, task tickets, or shared meeting memory. Ottex is strongest when meetings need to become finished written work such as follow-ups, CRM updates, tickets, docs, and PRDs. Granola wins for polished notes and team memory, while Otter.ai is better for teams that need shared workspaces, Salesforce or HubSpot sync, Slack or Zapier workflows, API/webhooks, and SOC 2 claims.
VoiceInk | Spokenly | Wispr Flow | ||||
|---|---|---|---|---|---|---|
| Decisions, owners, action itemsLeave the meeting with next steps clear | YesSource: ottex.ai — checked Jun 2026Decisions with owners extracted. | NoSource: dragon.nuance.com/en-us — checked Jun 2026Not supported. | NoSource: tryvoiceink.com/docs — checked Jun 2026Not supported. | NoSource: spokenly.app/docs — checked Jun 2026Not supported. | YesSource: help.granola.ai/article — checked Jun 2026Granola can extract decisions from meetings through templates and AI notes. | NoSource: wisprflow.ai/pricing — checked Jun 2026Not supported. |
| Follow-up email draftedThe outcome is sent, not summarized | YesSource: ottex.ai — checked Jun 2026Drafted from templates or your own. | NoSource: dragon.nuance.com/en-us — checked Jun 2026Not supported. | NoSource: tryvoiceink.com/docs — checked Jun 2026Not supported. | NoSource: spokenly.app/docs — checked Jun 2026Not supported. | YesSource: help.granola.ai/article — checked Jun 2026Recipes can generate follow-up emails from meetings. | NoSource: wisprflow.ai/pricing — checked Jun 2026Not supported. |
| CRM update / ticket draft | YesSource: ottex.ai — checked Jun 2026Structured for the target app. | NoSource: dragon.nuance.com/en-us — checked Jun 2026Not supported. | NoSource: tryvoiceink.com/docs — checked Jun 2026Not supported. | NoSource: spokenly.app/docs — checked Jun 2026Not supported. | YesSource: help.granola.ai/article — checked Jun 2026Granola lists Attio as an integration for sending meeting notes to CRM records. | NoSource: wisprflow.ai/pricing — checked Jun 2026Not supported. |
| Your rough notes steer the result | YesSource: ottex.ai — checked Jun 2026Rough notes and instructions steer the result. | NoSource: dragon.nuance.com/en-us — checked Jun 2026Not supported. | NoSource: tryvoiceink.com/docs — checked Jun 2026Not supported. | NoSource: spokenly.app/docs — checked Jun 2026Not supported. | YesSource: help.granola.ai/article — checked Jun 2026User notes are used to guide the generated meeting notes. | NoSource: wisprflow.ai/pricing — checked Jun 2026Not supported. |
Dragon has a mature Nuance Deep Learning speech engine, custom vocabulary, Auto-Texts, and voice commands, but it does not publish model selection, mixed-language speech, tone rewriting, translation, or BYOK. Ottex, Spokenly, MacWhisper, QuickWhisper, and VoiceInk all bring some mix of local models, BYOK, or one-time licensing, with Ottex and Spokenly both claiming 100+ languages. Wispr Flow and Willow Voice are polished cloud dictation options with 100+ languages, but their 2,000-word weekly free tiers are only about 15 minutes of speech.
VoiceInk | Spokenly | |||||
|---|---|---|---|---|---|---|
| Engines | Cloud: Whisper, Gemini 3 Flash, Voxtral. Local: Whisper, Parakeet v3.Source: ottex.ai — checked Jun 2026Hosted routing, BYOK, or fully local. | Nuance Deep Learning speech engineSource: dragon.nuance.com/en-us — checked Jun 2026Dragon Professional v16 uses a speech engine powered by Nuance Deep Learning technology. | Local models, Groq, ElevenLabs, Deepgram, Parakeet, Gemini, Mistral, Soniox, custom OpenAI-compatible modelsSource: tryvoiceink.com/docs — checked Jun 2026VoiceInk lists local, cloud, and custom transcription model options. | Whisper, Parakeet, Apple Speech Analyzer, GPT-4o Transcribe, Deepgram Nova 3, Soniox, Groq, ElevenLabs, Mistral, Fireworks, CerebrasSource: spokenly.app/pricing — checked Jun 2026Published pages name local, BYOK, managed cloud, and third-party providers across pricing and privacy materials. | No public claim | Whisper, Qwen 3 ASR, Nvidia Parakeet 110M, Nvidia Nemotron, Mistral Voxtral Mini, cloud providers OpenAI, Google, DeepGram, ElevenLabs, and GroqSource: quickwhisper.featurebase.app/changelog — checked Jun 2026Vendor changelog lists local voice model additions and cloud transcription providers. |
| Local / on-device optionAudio never leaves the machine | YesSource: ottex.ai — checked Jun 2026Whisper and Parakeet v3 run fully on-device. | PartialSource: dragon.nuance.com/en-us — checked Jun 2026Dragon Professional v16 is installed locally; Dragon Anywhere Mobile uses cloud dictation. | YesSource: tryvoiceink.com/docs — checked Jun 2026Local models run directly on the Mac without an internet connection. | YesSource: spokenly.app/docs — checked Jun 2026Local Only Mode keeps transcription on-device and blocks network access. | NoSource: help.granola.ai/article — checked Jun 2026Granola states it requires an internet connection to transcribe and generate notes. | YesSource: quickwhisper.app — checked Jun 2026Vendor says all transcriptions are processed and stored locally on the machine. |
| Model selection per task | YesSource: ottex.ai — checked Jun 2026Choose the model per task and per provider. | NoSource: dragon.nuance.com/en-us — checked Jun 2026No user-selectable recognition model is published. | YesSource: tryvoiceink.com/docs — checked Jun 2026Users can choose and set a default transcription model. | YesSource: spokenly.app/docs — checked Jun 2026Docs say users choose between fast and high-accuracy dictation models. | No public claim | YesSource: quickwhisper.featurebase.app/changelog — checked Jun 2026Vendor lists Whisper settings, custom GGML model import, and selectable local/cloud models. |
| Languages & mixed speechWork that crosses languages mid-sentence | YesSource: ottex.ai — checked Jun 2026100+ languages · Mid-sentence language switching. | No public claim | No public claim | No public claim | No public claim | No public claim |
| Per-app output rules | YesSource: ottex.ai — checked Jun 2026Per-app output rules: exact in AI chats, polished in email. | No public claim | YesSource: tryvoiceink.com/features — checked Jun 2026Power Mode can automatically apply app or website-specific transcription and enhancement settings. | No public claim | NoSource: granola.ai — checked Jun 2026Not supported | No public claim |
Ottex is unusually broad for a young Mac/iOS-first tool: it turns dictation and meetings into emails, tickets, CRM updates, notes, and docs, but it is not the pick if you need Windows, Android, web, browser extension, or compliance paperwork.
Personal or team — switch tabs. Subscriptions are flat; usage-based pricing depends on how much you speak and meet.
Aqua VoicePro Annual+ Granola for meetings — $168/yr$264/yr$96 + $168FAQ says Pro is $8/month billed annually after the free words. · Unlimited use
SpokenlyPro+ Granola for meetings — $168/yr$268/yr$100 + $168The pricing page shows annual billing at $99.99 per year and references Pro cloud features. · High-accuracy cloud models; AI text processing; works out of the box; priority support
Aqua VoicePro Monthly+ Granola for meetings — $168/yr$288/yr$120 + $168A vendor comparison page lists Pro at $10/month. · Unlimited use
Wispr FlowFlow Pro+ Granola for meetings — $168/yr$312/yr$144 + $168Annual price is published as $12/user/mo billed annually. · Unlimited words per week on Mac, Windows, iPhone, and Android
Wispr FlowFlow Pro+ Granola for meetings — $168/yr$348/yr$180 + $168Monthly price is published as $15/user/mo. · Unlimited words per week on Mac, Windows, iPhone, and AndroidEstimates, not quotes. Striped bar segments are the cost of the second app you picked above to cover the gap — rankings use that combined total. PAYG uses hosted at-cost estimates ($0.2/h dictation, $0.5/h meetings); BYOK is estimated at about a third of that because you pay the model provider directly. Bring-your-own-key plans still pay the model provider per use. One-time licenses are amortized over 2 years. Subscriptions use published pricing, checked Jun 2026.
The honest one-liner we'd give a friend for each. Select 2–4 and compare them side by side.
Best when: You want one tool for system-wide dictation, meeting recording, and finished written outcomes like follow-up emails, Linear tickets, CRM updates, PRDs, and docs
Best when: You want heavy-duty Windows dictation with custom words, Auto-Texts, and voice commands/macros.

Best when: You want native macOS dictation on Apple Silicon with local models that can run offline.

Best when: You want system-wide push-to-talk dictation on macOS, iOS, or Windows, with text inserted directly into the active app.
Best when: You want bot-free meeting capture that records from your mic and computer audio without joining calls.
Best when: You want local-first dictation, file transcription, and bot-free meeting recording on macOS 14.6+ with Apple Silicon.

Best when: You want system-wide voice dictation across Mac, Windows, iPhone, and Android.
Best when: You want offline-capable dictation and transcription on Mac or Windows with local Whisper/Parakeet plus optional cloud engines.

Best when: You want fast, system-wide voice dictation across Mac, Windows, or iOS for everyday writing.
Nothing breaks: both apps can run side by side while you decide.
Free out of the box: local models or the included starter credits, no card. Press the hotkey and speak — the default experience is deliberately close to Dragon NaturallySpeaking. Keep Dragon NaturallySpeaking installed; they coexist fine.
Set Ottex to the key you already press by muscle memory, or give it its own shortcut while you compare output side by side.
If Dragon NaturallySpeaking keeps a custom dictionary or replacement rules, copy those words into Ottex's personal dictionary and glossaries — names and jargon survive the move.
Defaults just work on starter credits. When you care, switch per task: local Whisper or Parakeet v3 for on-device privacy, your own key for control, or hosted routing at cost.
Set exact text for AI chats, polished email, structured tickets. Then record one low-stakes call and check the follow-up, CRM update, and ticket drafts.
Dragon NaturallySpeaking is Nuance's speech recognition product line for dictating text, using voice commands, adding custom vocabulary, and transcribing audio files. It fits best as a Windows-first dictation workhorse, not as a modern meeting assistant.
No public free plan is listed for Dragon Professional v16, Dragon Professional Anywhere, or Dragon Anywhere Mobile. If you need a real free path, Spokenly offers free unlimited local dictation, FluidVoice is free on macOS, and Ottex offers free local or BYOK options on Mac and iOS.
Nuance does not publish a public price for Dragon Professional v16, Dragon Professional Anywhere, or Dragon Anywhere Mobile in the provided facts. Cheaper transparent paths include VoiceInk at $25-$49 one-time, QuickWhisper Pro at $59 one-time, and Aqua Voice at $10/month or $96/year.
Yes. Dragon Professional v16 is optimized for Windows 11 and backwards-compatible with Windows 10, and Dragon Anywhere Mobile is available on Android in the US and Canada. If you want one dictation app across Mac, Windows, iOS, and Android, Wispr Flow is the broader replacement.
Partly. Dragon Professional v16 is installed locally for desktop dictation and audio-file transcription, but Dragon Anywhere Mobile uses cloud dictation. If offline or BYOK control is the reason you are switching, Spokenly, VoiceInk, QuickWhisper, MacWhisper, and Ottex are stronger candidates depending on platform.
Dragon Professional v16 has a local desktop story, but the facts do not show public HIPAA, SOC 2, GDPR, BYOK, or zero-data-retention claims for this product. Dragon Anywhere's terms say audio and text files are used on Nuance cloud servers to improve recognition, so privacy-focused buyers should look at local or BYOK tools such as Spokenly, VoiceInk, QuickWhisper, MacWhisper, or Ottex.