I’ve recently listened to the Useful AI MPU episode and thought I’d give Whisper Memos a go.
It seems decent but what does it give me that dictation in Drafts does not? Drafts does a great job of capturing my words and syncs everywhere almost instantly. It’s also under £2 a month. Whisper Memos is over £4 a month and only does dictation; it lacks all the other functionality of Drafts.
If anyone is wise in the ways of these apps I’d love to know the answer!
Am I missing something?
I’ve wondered that, too. I’ve been a Drafts user for 14 years, and admittedly, only use the dictation on the Apple Watch intermittently. I’m therefore not a prime candidate for these other services. I’ve always assumed they were for people comfortable with dictating a lot/for an extremely long time, perhaps? As I say, never ventured into them as it becomes “another thing to fiddle with.” But like you, intrigued nonetheless. Let’s hope others chime in.
I’m having a hard time justifying Drafts and Whisper Memos. rofl.
The only 3 subscriptions I have are:
FUTURE) Perplexity Pro or Claude or ChatGPT. I’ve got 2 years free Perplexity Pro, but in the future I can see myself paying for it as I use this all the time
These 3 subscriptions provide me incredible value. I try to find ways of stretching my dollars to the limit!
It gives you dictation parsed through the OpenAI Whisper model, and possibly a bit bit of a more direct startup to get into the dictation. Also, the actual recording is stored - not just the parsed version.
If the built-in Drafts dictation works for you, I can’t think of any reason to use WhisperMemos.
I’ve been trying Whisper Memos using the limited free transcriptions in the trial. Firstly it won’t replace Drafts - it complements it. Secondly, there are multiple transcription models to choose from, which do seem to provide better results than Drafts transcription. And thirdly, you can have it process the transcription with multiple custom prompts, so the one recording could for example result in a word for word transcription, a summary, a set of bullet points, a task list or whatever other format you might think up, all of which can be configured to be automatically emailed, or sent to one of the other integrations like Zapier. It seems to do a very good job at accurate transcription - better than the others I’ve tested.
It’s certainly quite powerful. Whether it’s worth the subscription will depend on your use case. If all you want is basic transcription, and you don’t mind a few mistakes and errors, then Drafts transcription should suffice. However, if you use dictation a lot, require a high degree of accuracy, or want the transcription processed further using custom prompts, then you’d be hard pressed to find something better than Whisper Memos.
I’m still undecided whether I can justify the subscription, but I am certainly tempted.
I mostly use the built-in transcription with Apple Notes, is Whisper Memo better than that?
For me, Whisper Memos is worth the subscription because of the Apple Watch complication. It’s one tap and start talking. Drafts still can’t do that, as far as I can tell.
This means that when I have something I need to remember, I can get it captured BEFORE I lose my train of thought and can’t remember what I wanted to remember!!
Like MacSparky, I’ve configured Whisper Memos to go to Drafts via email. I check my Drafts inbox every morning.
It’s OpenAI’s transcription, which is constantly evolving. When I used it, I generally found it better than Apple’s.
In case anyone is interested, prompted by this topic, I went on a little voyage of discovery to see if I could incorporate AI transcription along with some existing tools. Further, by not spending any more money along the way, too.
So this is what I came up with…
Yes, I know Just Press Record does transcription on iOS, but you have to get it out of the app itself. With the workflow above, I can now dictate at length on my Watch – without worrying about cut off times – and the AI transcription then appears in Drafts when I’m back at my Mac.
Edit - Took another look at this, and as drafts in Drafts can be created from outside of the app using AppleScript, I cut out the last 2 steps by having the Hazel script take the Whisper output and create a draft in Drafts direct. This means no waiting for Drafts to auto-import from the Inbox folder. If you have the MCP/CLI bits installed, you could just use the CLI I would think.
2nd edit - I was looking for a similarly quick way to do this on the Mac as on the Apple Watch. Just Press Record is available on the Mac, but I wanted something a bit quicker to invoke. I tried my copy of Piezo, but that involved having to name the file prior to creation. So, as is usually the case, in steps Sindre Sorhus with the free Recordia which does exactly what I was after. It sits in the menu bar with recordings started/stopped by a keyboard shortcut. Files are created and auto-named with time stamp. Hazel then does its bit to transcribe and create a new draft in Drafts just as per the other workflow.
Love the guide!
But out of curiosity, why do you use that specific record app vs something like just Apples Voice Memos? (Couldn’t that have given same/similar result?)
Just Press Record is an app I bought years ago. It records files direct to iCloud, meaning I know it’s easy for Hazel to act upon them. I don’t know anything about Apple’s Voice Memos. If it works in a similar manner, with no manual intervention required, then I see no reason why you couldn’t use that as an alternative.
I’ve pretty much settled on using SuperWhisper for all dictation. Yes, it’s a bit fiddly to set up, still working on it to be honest BUT single one time purchase, lots of choices for LLM models to help both on device and in the cloud, lots of options for adjusting and playing with setups to fine tune it Modes, which appears to be able to solve my you vs ewe problem I have with all other tools.
I’ve never clicked with drafts so that was never an option for me.