What's the best free voice memo transcription app for iPhone?

Voice memo transcription went from $10/month feature to free OS-level capability in September 2024 when Apple shipped iOS 18. For most iPhone users the answer is "the app already on your phone" — but five specific use cases still benefit from a third-party free option.

## Apple Voice Memos (iOS 18+) — the default winner

Free, on-device, no usage cap, works in 10 languages as of iOS 18.2. Transcription appears in the Transcript tab within 30 seconds of recording on iPhone XS or newer. Per <a href="https://9to5mac.com/whats-new-in-apple-notes-for-ios-18/">9to5Mac's iOS 18 Notes review</a>, the same recognizer is shared with the Notes audio recording feature and Voice Memos in iOS 18. Highest privacy guarantee since nothing leaves the device.

The three limits: requires iOS 18 + iPhone XS or newer; only 10 languages supported; recordings made before iOS 18 don't auto-transcribe without the trim-one-frame workaround.

## When to use a third-party free transcriber instead

• iPhone older than XS. No on-device transcription available. Use Otter or AssemblyAI free tier.
• Unsupported language (Arabic, Hindi, Dutch, Swedish, Vietnamese, etc.). Otter supports 40+ languages.
• Batch processing legacy memos. Apple has no batch retranscribe; importing into Némos or MacWhisper is faster.
• Speaker labels needed. Otter auto-labels multiple speakers; Apple Voice Memos produces one continuous transcript.
• Cross-platform sharing. Otter and Descript share transcripts to Android/web teammates.

## The free tier comparison

The five viable free options ranked by quality and usage cap:

• Apple Voice Memos. Free, unlimited, on-device. 10 languages. iPhone XS+ on iOS 18+.
• Otter.ai free tier. 600 minutes/month, 40 languages, speaker labels, web + iOS app. Cloud-based — audio uploaded.
• AssemblyAI free credits. $50 in free credits ~16 hours of audio. Highest accuracy on noisy audio per their published benchmarks.
• MacWhisper free tier. macOS only, but free for files under 5 minutes. Local Whisper model, no upload.
• Némos free tier. On-device transcription via Apple's Speech framework, unlimited audio files imported, no usage cap. The same engine as Apple Voice Memos with the iOS 18 gate removed.

## What "free" actually costs

Cloud transcribers — Otter, AssemblyAI, Descript free tier — upload your audio to their servers. Their terms of service grant them limited rights to process the audio for service operation. Otter's policy explicitly states transcripts are used to "improve our AI models" unless you opt out. For sensitive recordings (legal, medical, journalism source interviews), the "free" tier trades privacy for cost. On-device options — Apple Voice Memos, MacWhisper, Némos — do not upload, per Apple's <a href="https://developer.apple.com/documentation/Speech/SFSpeechRecognizer/supportsOnDeviceRecognition">SFSpeechRecognizer documentation</a> which guarantees no network access during on-device recognition.

## Recommendation by use case

• Personal voice notes on a recent iPhone: Apple Voice Memos. Nothing else needed.
• Old voice memos backlog: Némos free tier on iOS, or MacWhisper on Mac.
• Multi-speaker meeting recordings: Otter free tier (live transcription + speaker labels).
• Privacy-critical recordings: Apple Voice Memos or Némos. No cloud transcribers.
• Non-English language not supported by Apple: Otter or AssemblyAI.

## Bottom line

If you can update to iOS 18 and your iPhone is XS or newer, the best free transcriber is already installed. If not, Némos or Otter fill the gap, with Némos preferred when privacy matters and Otter preferred when speaker labels matter.