On device or in the cloud?
Both transcribe the same recording. The difference is not quality alone but what you accept in exchange: storage and waiting on one side, money and a connection on the other.
On the iPhone
Free, offline, unlimited.
- Apple's speech recognition is instantly there and needs no download
- Whisper models take 632 MB to 1.5 GB of storage
- One to three minutes of one-time preparation for the Neural Engine
- The model loads again after every app launch
- Nothing leaves the device
PhoneWhisperer Cloud
Seven days free, then billed by seconds.
- No download, no preparation, no loading after launch
- A model larger than any phone could hold
- You are charged once the transcript is actually back
- With no connection the recording waits — or runs locally with one tap
- The recording is transcribed in memory and discarded afterwards
The app tells you what your device can do
Settings → On this device says which models run on this exact iPhone — not as a table for every device, but for yours. An older iPhone with little memory cannot cope with the large Whisper models; the app says so instead of letting you find out the hard way. Apple's speech recognition runs either way.
What happens in the cloud
The recording passes through memory and is gone after the reply. It never touches a disk, the transcript is stored nowhere, and the log holds a timestamp, a duration and a status code — no text, no file name, no IP address. There is no account: the App Store knows who paid.