Helumi has a useful idea: let a Mac app open the piece of a big file it needs, without making you wait for the whole download. If you’re trying to play a video, why fetch the ending before you can watch the beginning?
I can see the appeal. I don’t think that makes it a reinvention of cloud storage, though. Google Drive already streams files, and Helumi’s own FAQ acknowledges that.
I haven’t tested the Mac app. What I checked was the website, the prices and the promises someone would be relying on before moving their files there. These notes are from 3 November 2026.
The useful question is how well it works
Streaming still downloads data, just as you need it. Think of bringing a few pages home from a library. The book hasn’t become smaller; the library still keeps the whole thing.
Google explains its streaming and caching behaviour. Apple has Optimize Mac Storage, and Dropbox has online-only files. They don’t all promise the same partial-file behaviour, but keeping a permanent copy of every cloud file on your disk is already optional.
I’d want to open a real editing project, scrub through a large video and see what happens when the connection drops. That would tell me more than calling streaming new. Even Google warns that video editing and heavy writing can be faster with mirrored files. An app that scans the whole file can eventually fetch all of it, however quick the first few seconds looked.
I’d be paying extra for that workflow
Here’s the 2 TB comparison. These are advertised USD monthly prices, without an annual commitment. Taxes and regional offers can change the checkout price.
| Service | Storage | Monthly price | Price per advertised TB |
|---|---|---|---|
| Helumi | 2 TB | $20 | $10 |
| iCloud+ | 2 TB | $9.99 | About $5 |
| Google One | 2 TB | $9.99 | About $5 |
| Dropbox Plus | 2 TB | $11.99 | About $6 |
Helumi’s slider starts at 500 GB for $5, offers 1 TB for $10, and goes up to 30 TB for $300. At 2 TB, it’s about twice Apple’s or Google’s monthly price. Both of those also offer family sharing.
If Helumi makes that work noticeably easier, the extra money could be worth it. I’d want to test that with my own files before moving an archive.
The storage bill raises another question. Helumi says it uses Cloudflare R2. R2 Standard’s public rate is $0.015 per GB-month. Keeping 1,000 GB there for a month would cost roughly $15 before operations and other costs. Helumi sells that capacity for $10.
I don’t know its actual bill. People leave quotas unused, storage classes differ, and negotiated rates could change the maths. The public numbers are a reason to ask how the service is funded, not proof that the business can’t work.
These promises should agree with each other
Its security page describes files encrypted on your Mac, encrypted filenames, recovery keys and encryption in 1 MiB pieces. Those pieces help explain how partial reads could work with encryption. There’s engineering there, although I haven’t independently checked the implementation.
The pages don’t quite agree, though. The homepage says “0 bytes,” while the privacy policy describes keeping recently opened pieces locally. The security page says filenames are hidden; the privacy policy says it keeps names and file details. I’d want that explained.
The payment promise matters even more. The homepage says files won’t be deleted when you stop paying. The terms allow removal from unpaid accounts after at least 30 days’ notice. They also describe early access, a three-day first-payment refund window and 90 days to export your files before a shutdown.
If I’m putting an archive there, which promise applies? Being a small company isn’t the problem. I’d want clear terms and working recovery and export with any provider.
So I wouldn’t dismiss Helumi as having done nothing. A good Mac drive with encrypted partial reads could be useful. It needs macOS 26.2, depends on the network for files you haven’t kept offline, and charges a premium. I’d try a copy of a real project, check offline work and recovery, then export the whole thing again before relying on it.
I also wanted to check whether better compression could shrink the files themselves. That’s a different problem, and the compression experiment produced some useful tradeoffs. It doesn’t benchmark Helumi, but it does show why saving space is more complicated than picking the algorithm with the smallest result.