Reviews 13 min read

Best Dictation App for iPhone: 7 Compared (2026)

The best dictation app for iPhone depends on the job: for typing by voice, start with Apple Dictation — it is free and runs on-device. For turning recordings into usable notes with summaries and secure sharing, Audio Notes AI is built for that job. For AI-polished live dictation anywhere on the phone, Wispr Flow. Full reasoning below.

On a computer? You can also convert audio to text in your browser with a free account.

We compared seven dictation, voice-to-text, and speech-to-text apps for iPhone on the same criteria: whether they work offline, what they do with your data, language coverage (including Cantonese), and price per hour of audio. Every price in this article was re-verified in September 2026 from each vendor's pricing page or App Store listing. We did not run our own accuracy benchmark, so this guide does not rank apps by accuracy. If your job is recorded audio rather than live typing, this doubles as our guide to the best transcription app for iPhone — and the best speech to text app overall for students, journalists, and professionals who record on iPhone.

The 7 Best Dictation Apps for iPhone at a Glance

App Best for Price (verified Sep 2026) How it transcribes Offline Keyboard
Apple Dictation / Voice Memos Free built-in baseline Free (in iOS) On-device (Apple engine) Yes (dictation) Built-in
Audio Notes AI Lectures & meetings + encrypted sharing Free 60 min lifetime; Pro free to try Cloud speech AI, summaries No No
Wispr Flow AI-polished live dictation Free 2,000 words/wk; Pro $15/mo ($12 annual) Cloud, rewrites into clean prose No Yes
Just Press Record One-time purchase, Watch capture $6.99 once Apple speech framework Partial No
Otter.ai Team meeting archives Free 300 min/mo; Pro $16.99/mo ($8.33 annual) Cloud, speaker labels No No
Notta Multilingual meetings & translation Free 120 min/mo; Pro $13.99/mo ($8.17 annual) Cloud, 58 languages No No
Aiko Fully offline file transcription ~$24 once On-device Whisper model Yes No

Dictation vs Transcription: Which Job Do You Have?

These terms get used interchangeably, but they are different jobs, and picking the wrong category wastes money:

  • Dictation is typing by voice. You speak, text appears where your cursor is — in Messages, Mail, WhatsApp, Notes. It is live, real-time input. Apple Dictation, Wispr Flow, and Superwhisper do this.
  • Transcription converts audio you already have — a lecture recording, a meeting, a voice memo — into text afterward. It is batch processing of files. Audio Notes AI, Otter, Notta, and Aiko do this.
  • Speech to text is the umbrella term covering both.

If your job is "write this email without typing," you want a dictation tool. If your job is "give me readable notes from the meeting I just recorded," you want a transcription app. Several apps below straddle both; each review says which side it favors.

How We Compared (Prices Verified September 2026)

Prices come from each vendor's pricing page or App Store listing, checked in September 2026. Offline behavior, data handling and language lists come from the vendors' own documentation and App Store privacy labels; where a claim comes from documentation rather than our own use, the review says so. We have not run a controlled accuracy benchmark across these apps, so there are no accuracy scores in this guide.

Disclosure: Audio Notes AI is our app. We position it where it fits — recorded lectures and meetings, imports, and encrypted sharing — and recommend competitors where they win, which is often.

Which Speech-to-Text App Is Most Accurate?

We have not run an independent accuracy benchmark, so we do not rank these apps by accuracy. On clear speech, the cloud apps (Audio Notes AI, Otter, Notta, Wispr Flow) and Apple Dictation all do well; noisy rooms and people talking over each other are hard for every app. The fair test is your own audio: try a real recording in the free tier of each app before you pay.

What matters in practice:

  • Clean audio: every cloud app here is good enough — pick on price, languages, and workflow.
  • Noisy, multi-speaker audio: hard for every app. Record close to the speaker, and test your own kind of recording in each free tier.
  • Live dictation into other apps: raw accuracy is not the whole story — Wispr Flow rewrites what you say into clean prose, which is why its output often needs the least editing.
  • Fully offline: Aiko runs smaller on-device Whisper models, trading some accuracy for zero network transmission — the right trade for confidential recordings.

The 7 Dictation and Voice-to-Text Apps, Reviewed

Apple Dictation & Voice Memos — Best Free Built-In Option

Every iPhone ships with two relevant tools. Dictation is the microphone button on the iOS keyboard: it types live speech into any text field, processed on-device in dozens of languages, so it works without internet. Notably, dictation supports Cantonese — both Hong Kong and China mainland variants — alongside Mandarin, Spanish, French, Japanese, and more (Apple's feature availability list).

  • Pros: Free forever, zero setup, on-device privacy, works offline, Cantonese dictation supported.
  • Cons: Voice Memos transcription covers exactly ten languages (English, Spanish, Portuguese, Italian, French, German, Japanese, Korean, Simplified and Traditional Chinese) — Cantonese is not one of them. No summaries, no speaker labels, no punctuation commands, transcripts stay locked inside the app.
  • Best for: short messages and quick thoughts where free and private beat everything else.

Audio Notes AI — Best for Lectures, Meetings, and Secure Sharing

Audio Notes AI records up to 90 minutes per session, transcribes server-side with speech AI, and automatically generates summaries, action items, and translations when the detected language differs from your app language. It transcribes 60+ languages with Cantonese treated as a first-class choice — one of the few iPhone speech-to-text apps that does.

  • Pros: Summaries, action items and translation come with every transcript. Imports audio and video files (mp3, m4a, wav, mp4, mov). Accepts voice messages directly from the WhatsApp/LINE share sheet. Shared notes are encrypted on-device with AES-256 before upload — the decryption key lives only in the link, and links expire after 24 hours.
  • Cons: Needs internet (cloud transcription). No speaker labels — transcripts come back as a single block. No keyboard, so it cannot type live into other apps. Free tier is 60 lifetime minutes, not monthly.
  • Pricing: Free for 60 minutes lifetime; Pro free to try with 10,000 minutes/month, iCloud sync and Siri Shortcuts.
  • Best for: students and professionals who record long sessions and want structured notes they can share securely. Deep dives: vs Apple Voice Memos, vs Otter.

Wispr Flow — Best AI-Polished Live Dictation

Wispr Flow is a dictation keyboard for iPhone (plus Mac and Windows) that does not just transcribe — it cleans up. Filler words removed, disfluent sentences rewritten into professional prose, delivered straight into whatever app your cursor is in. That cleanup pipeline is why writers and email-heavy users tolerate its cloud-only design.

  • Pros: Genuine keyboard extension that types into any app; excellent prose cleanup; cross-platform.
  • Cons: Transcription always runs in the cloud per the vendor's data-controls page — no offline mode. The free tier caps you at 2,000 words/week on desktop and 1,000 words/week on iPhone.
  • Pricing: Free tier as above; Pro $15/month, or $12/month billed annually (wisprflow.ai/pricing, verified Aug 2026).
  • Best for: people who dictate emails and documents all day and want publish-ready text, not raw transcripts.

Just Press Record — Best One-Time Purchase

Just Press Record is the anti-subscription: $6.99 once, covering iPhone, iPad, Mac, Apple Watch, and Vision Pro as a universal purchase. One tap records; transcripts generate using Apple's speech framework in 30+ languages; recordings can stay local or sync via your own iCloud Drive. Its App Store privacy label reads "Data Not Collected."

  • Pros: No subscription ever; Apple Watch recording; data stays yours; offline recording anywhere.
  • Cons: Transcription quality equals Apple's engine — solid for clear speech in its 30+ supported languages, but no AI summaries, no translation pipeline, and accuracy drops without a downloaded language pack.
  • Pricing: $6.99 one-time (App Store).
  • Best for: Watch-first capture and buyers who refuse subscriptions on principle.

Otter.ai — Best for Team Meeting Archives

Otter remains the default for organizations: it joins Zoom, Meet, and Teams calls as a bot, produces searchable archives with speaker identification, and integrates with calendars and CRMs. Its speaker labels are the most complete of any app here.

  • Pros: Speaker detection; meeting-bot automation; strong search across an entire archive; mature sharing.
  • Cons: Cloud-only. Language support is six: English, Spanish, French, German, Japanese, and Chinese (Simplified) — no Cantonese, per Otter's own help center. Your transcripts train its models unless you opt out on paid tiers.
  • Pricing: Free 300 minutes/month (30-minute conversation cap); Pro $16.99/month or $8.33/month billed annually; Business $30/month or $19.99/month annually (otter.ai/pricing, verified Aug 2026).
  • Best for: teams who live in scheduled meetings. Full breakdown: Audio Notes vs Otter.

Notta — Best Multilingual Meetings and Translation

Notta covers 58 languages — the widest verified set of the cloud tools here — with real-time translation between them, a meeting bot for Zoom/Meet/Teams, and a dedicated Cantonese speech-to-text tool on its site, which most competitors cannot say.

  • Pros: Broadest language list including Cantonese; translation built in; SOC 2 and ISO 27001 certified.
  • Cons: The free plan caps conversations at three minutes, which rules out real meetings; lower tiers may use your data for model training unless you opt out on Enterprise; cloud-only.
  • Pricing: Free 120 minutes/month; Pro $13.99/month or $8.17/month billed annually; Business from $27.99/month (notta.ai/en/pricing, verified Aug 2026).
  • Best for: multilingual teams whose meetings switch languages. See also Audio Notes vs Notta.

Aiko — Best Fully Offline Transcription

Aiko runs OpenAI's Whisper locally on your device via whisper.cpp — nothing leaves the phone, and it works in airplane mode. It is built for files you already have rather than live dictation: import audio, get a transcript, in roughly 100 languages. The developer, Sindre Sorhus, sells it as a one-time purchase (~$24 universal) with no subscription (sindresorhus.com/aiko).

  • Pros: True on-device privacy; airplane-mode proof; no ongoing cost.
  • Cons: On-device models are smaller than server-side speech AI, so accuracy trails cloud tools; no summaries, translation pipeline, or sharing; import-only workflow.
  • Honorable mention: Superwhisper ($8.49/month, $84.99/year, or $249.99 lifetime) adds a dictation keyboard with offline support — but only with its two local models, English-only on iOS; everything else requires internet.
  • Best for: journalists, lawyers, and anyone whose threat model demands zero network transmission. More options: Audio Notes vs DIY Whisper apps.

Note: Dragon Anywhere, long the premium dictation name, stopped new subscriptions and renewals on July 1, 2026, per its store listings — we no longer recommend it for new buyers.

Privacy and Offline: Which Apps Work in Airplane Mode?

Cloud transcription is accurate, but some recordings — legal, medical, personal — should never leave your device. Here is the matrix:

App Works offline? Where audio is processed
Apple Dictation Yes (downloaded languages) On-device
Aiko Yes On-device (Whisper via whisper.cpp)
Superwhisper Partial — local models only, English on iOS On-device or your chosen cloud provider
Just Press Record Partial — recording yes; transcript needs language pack On-device (Apple framework)
Audio Notes AI No Cloud (speech AI); shares encrypted client-side before upload
Wispr Flow No Cloud (always)
Otter.ai No Cloud
Notta No Cloud

One nuance: "cloud" is not automatically unsafe. Audio Notes AI encrypts shared notes on-device with AES-256 before anything uploads — the server stores ciphertext it cannot read, and links self-destruct after 24 hours. But for raw recording privacy, on-device tools are categorically different: with Apple Dictation or Aiko, no network transmission happens at all.

Cantonese and Multilingual Accuracy

Cantonese deserves its own section because mainstream tools keep failing it. The verified picture as of August 2026:

  • Apple Dictation: supports Cantonese (Hong Kong and China mainland), on-device — good for live typing, but it produces no saved transcript.
  • Apple Voice Memos: ten transcription languages, no Cantonese.
  • Otter.ai: six languages, no Cantonese.
  • Notta: supports Cantonese with a dedicated tool page.
  • AI-based apps (Audio Notes AI, Aiko): Modern speech models support Cantonese as its own language. Expect strong results on clear speech, with mixed Cantonese-English chatter still producing occasional errors.

If Cantonese is your primary language, our dedicated guides cover the workflows in depth: transcribing Cantonese audio to text (English) and the Traditional-Chinese guide to 廣東話語音轉文字. Broader language coverage lives in the any-language transcription hub.

What Dictation Costs per Hour of Audio (September 2026)

For live iPhone dictation — typing by voice into Messages or Mail — the marginal cost is zero on every app here: Apple Dictation is free, and Wispr Flow free tier covers 1,000 words a week on iPhone. The money question only bites for transcription of recorded speech-to-text audio, where pricing models diverge hard:

  • Free built-ins: Apple Dictation and Voice Memos cost nothing but give you a bare transcript with no summary, translation, or export.
  • Flat-rate AI plans: ; Otter Pro at $16.99/month for 1,200 minutes is about 1.4 cents per minute; Notta Pro at $13.99/month sits between them with translation as a paid add-on.
  • Human transcription: Rev human service at $1.99 per minute means one hour of audio costs $119.40 — more than a year of Audio Notes Pro. Worth it for legal or verbatim work; absurd for lecture notes.
  • One-time purchases: Just Press Record ($6.99 once) and Aiko (~$24 once) cost nothing per minute ever, but neither writes summaries nor translates — you trade ongoing cost for ongoing manual work.

Rule of thumb: if you transcribe more than two hours a month, a flat-rate plan beats free-tier hopping; if you transcribe less, the 60 lifetime minutes in Audio Notes free tier plus Apple built-ins cover you without a subscription.

What Audio Notes AI Does Differently (and What It Does Not Do)

The apps above split into three philosophies, and Audio Notes occupies exactly one of them. It is a recording-first transcription app: you record up to 90 minutes or import a file you already own (a voice memo, a lecture video, a WhatsApp voice message from the share sheet), and you get back a transcript plus an AI summary, action items, and translation — with a share link that is encrypted on your phone before upload and dies after 24 hours.

What it does not do, so you do not buy it for the wrong job:

  • No live keyboard dictation — it cannot type into other apps as you speak. That is Wispr Flow and Apple Dictation territory.
  • No meeting bot — it never joins your Zoom or Teams call. Otter and Notta own that workflow.
  • No speaker labels — multi-speaker transcripts come back as a single block of text.
  • No offline mode — transcription runs on cloud AI, so airplane mode is a non-starter (Aiko is the offline answer).

The reason it still wins the lecture-and-meeting-recording job on iPhone: nothing else combines state-of-the-art accuracy on noisy audio, 60+ languages with explicit Cantonese support, video file import, automatic translation and summaries, and zero-knowledge encrypted sharing, free to try. Each of those exists elsewhere; the combination does not.

The Verdict: Which One Should You Use?

Choose Apple Dictation if you want free and private above all

Quick messages, short notes, zero setup, works offline — and it is already installed. Start here before paying anyone.

Choose Audio Notes AI if you record lectures, meetings, or voice memos you will reuse

You get state-of-the-art accuracy, automatic summaries and action items, 60+ languages with explicit Cantonese support, video import, WhatsApp share-sheet intake, and encrypted expiring links, free to try with 60 lifetime minutes included.

Choose Wispr Flow if you dictate documents all day

Its keyboard types polished prose into every app on your phone. Accept the cloud requirement and word limits.

Choose Just Press Record or Aiko if subscriptions offend you or privacy is absolute

$6.99 buys a lifetime of Watch-friendly recording; ~$24 buys fully offline Whisper transcription of existing files.

Choose Otter or Notta for team meeting infrastructure

Speaker-labeled, bot-joined, searchable archives — Otter for English-first teams, Notta when your meetings span many languages.

The bottom line: there is no single best app, because dictation and transcription are different jobs — and the best voice to text app for iPhone is whichever matches yours. Match the tool to the job, verify prices before subscribing (they change), and prefer tools that tell you where your audio goes.

Frequently Asked Questions

What is the best dictation app for iPhone in 2026?

For most people, Apple's built-in Dictation is the best starting point: free, on-device, and it supports Cantonese among 40+ dictation languages. For recorded meetings and lectures that need transcripts, summaries, and secure sharing, Audio Notes AI is our pick — AI-powered, 60+ languages including Cantonese, free for 60 lifetime minutes. Wispr Flow is strongest for live AI-polished dictation into any app via its keyboard.

What is the best free voice to text app for iPhone?

Apple Dictation and Voice Memos are genuinely free and unlimited-in-principle because they run on your device. If you want AI summaries and action items too, Audio Notes AI includes 60 minutes of lifetime transcription free — a one-time allowance rather than a subscription trial.

Which iPhone dictation apps work offline in airplane mode?

Apple Dictation works offline for downloaded languages because processing happens on-device. Aiko transcribes entirely on-device with Whisper and works in airplane mode. Superwhisper works offline only with its two local models (English-only on iOS). Cloud tools — Otter, Notta, Wispr Flow, and Audio Notes AI — need an internet connection.

What is the difference between dictation and transcription?

Dictation converts live speech into text as you speak — typing by voice into any text field. Transcription converts a recording you already have — a meeting, lecture, or voice memo — into text afterward. Dictation is real-time input; transcription is post-processing of files.

Does the iPhone support Cantonese voice to text?

Yes for dictation: iOS keyboard Dictation supports Cantonese (Hong Kong) and Cantonese (China mainland), processed on-device. No for Voice Memos: its transcription feature covers ten languages and none of them is Cantonese. Third-party apps fill the gap — Audio Notes AI lists Cantonese explicitly in its language picker, and Notta offers a dedicated Cantonese tool.

Which app is best for transcribing WhatsApp voice messages?

WhatsApp's built-in transcript (November 2024 onward) is on-device but view-only — you cannot copy or export it. To get copyable, searchable text, forward the message to a transcription app via the share sheet. Our WhatsApp voice message guide walks through both routes step by step.

Which speech to text app is most accurate on iPhone?

We have not run an independent accuracy benchmark, so we do not rank these apps by accuracy. On clear speech, the cloud apps (Audio Notes AI, Otter, Notta, Wispr Flow) and Apple Dictation all do well; noisy rooms and people talking over each other are hard for every app. The fair test is your own audio: try a real recording in the free tier of each app before you pay.

How much does iPhone dictation cost per hour of audio?

Apple Dictation and Voice Memos are free. Among paid tools, Audio Notes is free to try, while human services like Rev charge $1.99 per minute — one hour costs $119.40. If you transcribe regularly, a flat-rate AI plan beats per-minute pricing by two orders of magnitude.

Can ChatGPT transcribe audio on iPhone?

ChatGPT on iPhone is built for conversation and text. To turn a recording you already own — a voice memo, an MP4, a WhatsApp audio — into text with translation and summary, import it into Audio Notes AI via the share sheet: speech AI transcribes server-side in 60+ languages in about 30 seconds. The step-by-step for message audio is in our voice-message transcription guide.

Start Turning Talk Into Notes

Download Audio Notes AI free on the App Store — 60 minutes of lifetime transcription included — and compare it against whatever you are using today. If you share notes, see how encrypted voice memo sharing works.

Transcribe & Secure Your Audio

Record voice notes, generate instant summaries, and share them via secure expiring links with zero-knowledge encryption.

Download on App Store
Download on App Store