Hey! Blip co-founder here. We didn't expect to show up on HN, but really grateful to OP for sharing Blip. Here's a little bit more about it.
We've built Blip because it's still hard to send original quality photos, videos, and large files to your devices and to other people on the internet. We’re designers and engineers, so our goal has always been to keep the product super simple on the surface, but really fast and powerful underneath.
Blip works in a peer-to-peer way at the UI level: you pick the device or person, and Blip takes care of the delivery. Transfers go directly over WAN whenever possible, and fall back to relays when needed. The idea is to send in one click, skipping the usual dance of moving files through cloud drives and managing shared links.
Under the hood, Blip is optimized for large media and data transfers. It supports full-speed acceleration, resumable progress, and we're rolling out E2EE across all clients to ensure sensitive business data remains secure. Many creative pros and teams already use Blip in their daily media workflows.
We don’t monetize data because it doesn't align with the values of our creative and technical users. Instead, we run on a simple donation and subscription model that lets you support the product and use it without limits, quotas, and frustrations. Our goal is to make file transfer feel invisible.
Thank you, and great question. In my experience, big companies have way more strategic priorities than two guys who just want to build something useful. We didn't leave to build exactly this, but a few things came together organically from past projects, including our work at Dropbox.
A compromised relay server can't access the data because it's encrypted.
A meaningful vulnerability would have to be in either the software itself or in the coordination server. That attack surface is the same whether or not you have relays.
You can reduce the attack surface to just the software if there's a way for users to verify keys manually. But again, same attack surface whether or not you have relays.
That's right. We actively manage load across our relay network to ensure good performance, but we'll prioritize business transfers during peak times. We don't artificially limit the client, but P2P connection speeds can sometimes be affected by router configurations and ISP routing. For example, some ISPs route P2P traffic through slower paths, which can introduce variability.
We evaluated QUIC (and many other approaches). Turns out it's a lot harder than you might think to move traffic at high speed across the world, over residential-grade internet, and not drain your battery.
If it's truly p2p, some relay would be there in case the client cannot be reached through NAT. Not sure how they would bear the cost of the bandwidth for unlimited transfers in that case
Syncing files is only one of its capabilities. It is best suited for use between two individuals/organizations that need to transfer files between themselves more than a few times. Its main advantage for me is that it doesn't require a central storage system, such as Dropbox, to hold the files. It's just you and me and a rendezvous server.
In a previous life, I used Syncthing to transfer terabytes of files from the company I worked for to a third-party printer. It was delightfully reliable, and easy to set up.
If it's a one-off, yeah, you're right that it's not wonderful. I've used Magic Wormhole successfully for that use case.
I wonder how widely usable is file sharing nowadays when most of the non tech people just use cloud services for their data, be it google docs or some cloud photo storage
Most non-tech people do not just use cloud services for their data.
Really not sure where you got that from, but even if it was true, most non-tech people will still shy away from putting a 250G file in a cloud service once they get prompted to upgrade their plan because they don't have enough space.
Right. We don't provide storage. Blip is designed to be the fastest way to send things to your devices and to other people in real time, without waiting for uploads, sync, or managing shared links.
I dunno, seems like most do? Their stuff is in Google Photos, Google Docs, iCloud Drive, etc. And yes they pay for the space once their photos or phone backups get big enough.
And I don't know any non-tech people who have any 250GB files. The only people I know with those shoot 4K video professionally. Or scientists running truly massive simulations.
I think non-tech people used to be able to, but tech companies have been on a 10+ year long crusade against the concept of a "file" and where that file is "stored" and trying to blur once-sharp lines so that people forget. Tech really wants you to think of your data as an amorphous blob vaguely "in their app" and not worry about crisp delineations like files, whose hard drive those files are on, and whose machine that hard drive is in.
Who else needs to share files bigger than 1 TB? Most cloud services, such as Google Drive or OneDrive, are more than sufficient for managing massive files. I don't see the appeal of this new service.
Oddly enough, I would argue the exact opposite direction; really big files are exactly where I want to do a direct p2p transfer without paying to store it in the cloud.
Teams have been building services like this for ~20 years. They very rarely stick. I agree that it’s still too hard, but the market has historically been too small. It’s at-best a hobby side project for a larger company that can afford to burn the cash.
Like Apple offers this for email for iCloud users. I think Firefox Relay offered it too. There was another company in the late 00s that offered a P2P version.
I mean _cool_, but I’ve not seen a company with this as its primary product last more than 18 months before. With that, _good luck_.
> How is Blip different to nearby sharing like AirDrop?
Apple’s “AirDrop” and Google’s “Nearby Share” can be really handy. However, they aren’t compatible with each other and require devices to be physically next to each other. They are also unreliable when transferring large files, and will often lose your progress.
> Blip doesn’t need devices to be nearby, so it’s much more reliable. Blip works wherever your internet connected devices are in the world, and works regardless of what kind of device you own. You can transfer from Android to Mac, Windows to iPhone, iPad to Android—you name it!
Cheap has many more meanings than that. It can mean comparatively inexpensive (a cheap Lamborghini), of inferior quality (cheap paperclip), miserly (he's too cheap to buy better), gained with little effort (cheap win).
Surprised syncthing isn't mentioned yet. It has been the most stable sync tool for me over the years https://syncthing.net/ . Solid product . Great oboarding experience for Blip! Its just working!
This looks really cool. I especially like the "Keep your progress, whatever happens" feature.
The product looks polished and I definitely see myself using it. My only concern is: Are they taking VC money? Is this going to be enshittified to death trying to pursue a 1000x investment return?
1. For a Linux user, you can already build such a system yourself quite trivially by getting an FTP account, mounting it locally with curlftpfs, and then using SVN or CVS on the mounted filesystem. From Windows or Mac, this FTP account could be accessed through built-in software.
2. It doesn't actually replace a USB drive. Most people I know e-mail files to themselves or host them somewhere online to be able to perform presentations, but they still carry a USB drive in case there are connectivity problems. This does not solve the connectivity issue.
3. It does not seem very "viral" or income-generating. I know this is premature at this point, but without charging users for the service, is it reasonable to expect to make money off of this?
> you can already build such a system yourself quite trivially by getting an FTP account, mounting it locally with curlftpfs, and then using SVN or CVS on the mounted filesystem...
I don't think the word "trivially" means what you think it means.
[Edit: I now realize the above is a verbatim quote from a naysayer after the Dropbox announcement]
"former dropbox engineers" doesn't mean a whole lot -- it's a large company where tens of thousands of people have worked over the years. It's not like this is by the founders or anything.
Yea, "former [COMPANY] employee" could mean anything. I'm not sure it's really much of a flex. I'm a former Apple employee. Nobody gives a shit, and nobody should--That doesn't count for anything if I were to do a software startup. It wouldn't even bear mentioning in a press release.
So you run the app. The other person runs the app.
How do those 2 programs find out about each other?
Well, they need to use some server that will connect all users, allow you to find the other person in order to send them a file.
Even BitTorrent needs tracker servers to connect downloaders to uploaders.
Now, you could try to make it run without a company behind it, like BitTorrent.
I guess it's an exercise for the reader.
Someone still needs to run those connecting servers out of the goodness of their heart.
You would need to degrade the usability (the first thing a user would need to do is to configure the app with the address of at least one connecting server).
BitTorrent is anonymous but for sending files you need to connect to a person you know. If you enable contacting via the protocol, it'll be a phishing nightmare.
So you'll have to degrade usability some more and use a side channel (e.g. an email) to exchange identities with people you want to send file to (or receive files from).
By having a service and a company behind it, they are able to perfect the UX and run the necessary servers to implement that UX.
Universities, the Debian Foundation, Nvidia, the ccc, hell even Microsoft run magic-wormhole relays... the only thing new here is the subscription and the hipster gui client.
This is classic critique. Dropbox was criticized in the same way when it was announced years ago.
Magic Wormhole is cool (I just learned about it! Thanks! :)), but most people don't use nor care about the command line, and are probably afraid of it.
I think that might be billed on the sender vs the receiver. So if you're a solo creative person sharing stuff with clients, this can be one way of doing that. Also potentially useful for small businesses without an enterprise file sharing solution.
Hey! Blip co-founder here. We didn't expect to show up on HN, but really grateful to OP for sharing Blip. Here's a little bit more about it.
We've built Blip because it's still hard to send original quality photos, videos, and large files to your devices and to other people on the internet. We’re designers and engineers, so our goal has always been to keep the product super simple on the surface, but really fast and powerful underneath.
Blip works in a peer-to-peer way at the UI level: you pick the device or person, and Blip takes care of the delivery. Transfers go directly over WAN whenever possible, and fall back to relays when needed. The idea is to send in one click, skipping the usual dance of moving files through cloud drives and managing shared links.
Under the hood, Blip is optimized for large media and data transfers. It supports full-speed acceleration, resumable progress, and we're rolling out E2EE across all clients to ensure sensitive business data remains secure. Many creative pros and teams already use Blip in their daily media workflows.
We don’t monetize data because it doesn't align with the values of our creative and technical users. Instead, we run on a simple donation and subscription model that lets you support the product and use it without limits, quotas, and frustrations. Our goal is to make file transfer feel invisible.
Happy to answer any questions.
Looks amazing! Maybe a dumb question: why isn't Dropbox doing this? Why did you all need to leave to make this a reality?
Thank you, and great question. In my experience, big companies have way more strategic priorities than two guys who just want to build something useful. We didn't leave to build exactly this, but a few things came together organically from past projects, including our work at Dropbox.
Linux support please!
We hear you :)
What's the timeline on rolling out E2EE? Is it for paid users only?
Gold standard on all plans. Already supported for transfers between iOS and Mac devices. Android and Windows coming soon.
Interesting that it says "Internet sending may be slower during peak times to keep things fair" even though it's supposed to be P2P?
Maybe they just mean if you end up with a relayed connection due to NAT issues? Because lower down it says "Send as fast as your connection"
> When a direct connection isn’t possible, files travel through our servers.
I figured, it's just explained in an odd way.
That's what actually made it a hard pass for me.
Why's that, given that files are encrypted?
With whose key? I am not saying the service is not trust worthy, just that there is trust involved.
Trust not only they are not malicious, but also they won't have some kind of vulnerability.
Plus if it's encrypted how is the other party going to read the file? The key will have to take the same path.
> Trust not only they are not malicious, but also they won't have some kind of vulnerability.
Wouldn't that still be the case if relay servers didn't exist? A hacked version can send your file to the wrong person.
There is more attack surface with a server.
The vulnerability doesn't even have to be in their software, but in any piece of software they use, ssh, nginx, etc.
A compromised relay server can't access the data because it's encrypted.
A meaningful vulnerability would have to be in either the software itself or in the coordination server. That attack surface is the same whether or not you have relays.
You can reduce the attack surface to just the software if there's a way for users to verify keys manually. But again, same attack surface whether or not you have relays.
Hopefully it's a privpub negotiation. But yes, you have to trust the code.
That's right. We actively manage load across our relay network to ensure good performance, but we'll prioritize business transfers during peak times. We don't artificially limit the client, but P2P connection speeds can sometimes be affected by router configurations and ISP routing. For example, some ISPs route P2P traffic through slower paths, which can introduce variability.
Does anyone have an idea of how this is built? I wonder if they are using QUIC with relay servers or something like Tailscale's DERP.
It's something closer to Tailscale DERP.
We evaluated QUIC (and many other approaches). Turns out it's a lot harder than you might think to move traffic at high speed across the world, over residential-grade internet, and not drain your battery.
Did you consider Iroh and if you chose not to use it, why not?
If it's truly p2p, some relay would be there in case the client cannot be reached through NAT. Not sure how they would bear the cost of the bandwidth for unlimited transfers in that case
Traffic is pretty cheap outside the big clouds. For example Hetzner charges $1/TB on a 10Gbit connection
Their 1Gbit connections are unmetered.
I wouldn't be surprised if it's built with Iroh
https://www.iroh.computer/
I wonder what tech stack they're using, given it supports all the major platforms. I like sendme [0] which uses iroh, a peer-to-peer library as well.
[0] https://github.com/n0-computer/sendme
Free open source alternative: https://pairdrop.net
Magic Wormhole [1] also exists.
[1] https://magic-wormhole.readthedocs.io/en/latest/
And for a web native version, there's also WebWormhole: https://webwormhole.com/
Syncthing [1] also exists
[1] https://syncthing.net/
Syncthing is for syncing folders like Dropbox
It's not a good solution to send individual large files.
Syncing files is only one of its capabilities. It is best suited for use between two individuals/organizations that need to transfer files between themselves more than a few times. Its main advantage for me is that it doesn't require a central storage system, such as Dropbox, to hold the files. It's just you and me and a rendezvous server.
In a previous life, I used Syncthing to transfer terabytes of files from the company I worked for to a third-party printer. It was delightfully reliable, and easy to set up.
If it's a one-off, yeah, you're right that it's not wonderful. I've used Magic Wormhole successfully for that use case.
So it’s the GPG of file sharing?
Is there something like a Magic Wormhole server, so I can e.g host a file on my NAS (behind a NAT) for download long term?
also Keet: https://keet.io/
There's also LocalSend, which I've found works the best for me personally and is a bit more polished than browser clients
There's also croc: https://github.com/schollz/croc
croc has the advantage of being well-established by now. Most package systems have it.
I vibe coded this in one hour to send files to my work laptop. Static page + webRTC + short lived cloudflare durable object to make the handshake.
https://send.drexl.dev/
..also FilePizza as a web service: https://github.com/kern/filepizza
or tailscale's Taildrop as a native application: https://tailscale.com/kb/1106/taildrop
Are you people seriously suggesting webrtc crap in response to a native app built for much much high speed transfers? Unbelievable
[dead]
This looks awesome. For sending files between my phone (Android) and my son's iPad, I use:
Android: Wormhole William (https://github.com/psanford/wormhole-william-mobile)
iOS: Destiny (https://github.com/LeastAuthority/destiny)
Some drawbacks to my current approach:
1. Destiny needs to be configured to use the standard Magic Wormhole servers (just once, after installation): https://github.com/LeastAuthority/destiny/issues/259#issueco...
2. Initiating a transfer requires out of band communication and some copy+paste.
I wonder how widely usable is file sharing nowadays when most of the non tech people just use cloud services for their data, be it google docs or some cloud photo storage
Most non-tech people do not just use cloud services for their data.
Really not sure where you got that from, but even if it was true, most non-tech people will still shy away from putting a 250G file in a cloud service once they get prompted to upgrade their plan because they don't have enough space.
Cloud is kinda the default now. Most Americans just take pictures with their iPhones and it ends up in icloud.
That's different from getting a Dropbox account or manually uploading stuff to Google drive in order to share it with someone.
Right. We don't provide storage. Blip is designed to be the fastest way to send things to your devices and to other people in real time, without waiting for uploads, sync, or managing shared links.
The point is your stuff is in there already. You just click share and send somebody a link.
I dunno, seems like most do? Their stuff is in Google Photos, Google Docs, iCloud Drive, etc. And yes they pay for the space once their photos or phone backups get big enough.
And I don't know any non-tech people who have any 250GB files. The only people I know with those shoot 4K video professionally. Or scientists running truly massive simulations.
Non-tech people couldn't tell you what a "file" is.
I think non-tech people used to be able to, but tech companies have been on a 10+ year long crusade against the concept of a "file" and where that file is "stored" and trying to blur once-sharp lines so that people forget. Tech really wants you to think of your data as an amorphous blob vaguely "in their app" and not worry about crisp delineations like files, whose hard drive those files are on, and whose machine that hard drive is in.
Who else needs to share files bigger than 1 TB? Most cloud services, such as Google Drive or OneDrive, are more than sufficient for managing massive files. I don't see the appeal of this new service.
Oddly enough, I would argue the exact opposite direction; really big files are exactly where I want to do a direct p2p transfer without paying to store it in the cloud.
Google Drive throttles uploads over 5GB, and not everyone has a storage plan that could fit that.
It would save you the time to upload first, effectively halving the time required.
Teams have been building services like this for ~20 years. They very rarely stick. I agree that it’s still too hard, but the market has historically been too small. It’s at-best a hobby side project for a larger company that can afford to burn the cash.
Like Apple offers this for email for iCloud users. I think Firefox Relay offered it too. There was another company in the late 00s that offered a P2P version.
I mean _cool_, but I’ve not seen a company with this as its primary product last more than 18 months before. With that, _good luck_.
Wait... you're telling me it's 2025 and we can finally conveniently send files?
Looks not really convenient...
I'd be more inclined to use this if it were open source. Oh well.
Android app looks beautiful. Waiting for the Linux version.
Thanks. We put a lot of work into our apps.
Is this AirDrop but cross-platform?
Addressed in the FAQ:
> How is Blip different to nearby sharing like AirDrop? Apple’s “AirDrop” and Google’s “Nearby Share” can be really handy. However, they aren’t compatible with each other and require devices to be physically next to each other. They are also unreliable when transferring large files, and will often lose your progress.
> Blip doesn’t need devices to be nearby, so it’s much more reliable. Blip works wherever your internet connected devices are in the world, and works regardless of what kind of device you own. You can transfer from Android to Mac, Windows to iPhone, iPad to Android—you name it!
No.
AirDrop is for people who are physically nearby.
This allows to send files between any computers anywhere.
The other person must be a known contact but it doesn't have to be on the same local network like in AirDrop.
There’s a new AirDrop feature that also transfers over the internet: https://allthings.how/how-airdrop-over-internet-works-on-iph...
> To start AirDrop over the Internet, initiate the AirDrop as you normally would
...
this is hardly the same thing?
This is a minor OCD nitpick, but
> at super fast speeds
Fast speeds aren't a thing, just like cheap prices and wet waters aren't.
I don’t follow. Speed and price can be many values, high or low. It’s perfectly valid to add an adjective describing it as fast or cheap.
Cheap means "low in price", so "cheap price" makes no sense.
Fast means "of a high speed", so it's the same here.
nit-nit, what's wrong with cheap prices? price doesnt infer magnitude.
Cheap means "low in price". Price can't be "low in price", it can be just "low".
Cheap has many more meanings than that. It can mean comparatively inexpensive (a cheap Lamborghini), of inferior quality (cheap paperclip), miserly (he's too cheap to buy better), gained with little effort (cheap win).
So what's a cheap price? A price of inferior quality? A miserly price? A price obtained with little effort? Or is after all a low price what's meant?
Sure. Still doesn't make "cheap price" a valid combo.
could be a contraction of cheap at the price https://en.m.wiktionary.org/wiki/cheap_at_the_price
That's a quality nit pick for sure.
J/k
It's a high quality nit pick imo:)
Somehow it has to be contrasted with ``slow'' speeds.
"High speeds"
It's like YouSendIt!
Surprised syncthing isn't mentioned yet. It has been the most stable sync tool for me over the years https://syncthing.net/ . Solid product . Great oboarding experience for Blip! Its just working!
This looks really cool. I especially like the "Keep your progress, whatever happens" feature.
The product looks polished and I definitely see myself using it. My only concern is: Are they taking VC money? Is this going to be enshittified to death trying to pursue a 1000x investment return?
No benchmark comparisons with Aspera or similar even though they compare pricing? Because my gut instinct is it's probably a good bit slower.
That's not what our professional customers are telling us :) Good point about adding benchmarks to our website though.
That's awesome man, I'm glad to hear there's a viable competitor. Gl and I look forward to seeing it.
Don't see any mention of an API.
We don't have the API yet, but are exploring use cases for it. Send me an email at tomaz@blip.net -- would love to chat.
I have a few qualms with this app:
1. For a Linux user, you can already build such a system yourself quite trivially by getting an FTP account, mounting it locally with curlftpfs, and then using SVN or CVS on the mounted filesystem. From Windows or Mac, this FTP account could be accessed through built-in software.
2. It doesn't actually replace a USB drive. Most people I know e-mail files to themselves or host them somewhere online to be able to perform presentations, but they still carry a USB drive in case there are connectivity problems. This does not solve the connectivity issue.
3. It does not seem very "viral" or income-generating. I know this is premature at this point, but without charging users for the service, is it reasonable to expect to make money off of this?
Context, for the above: https://news.ycombinator.com/item?id=9224
It worked out the first time!
Thanks for this - I thought the OP was insane...
Hey isn't this the same famous reply as the one the Dropbox founder got here in HN ?
infamous*
but yes, sure sounds like it!
1. - FTP could fail / doesn't resume the same way rsync, mutagen, or syncthing could
2. - nothing would by your criteria of an airgap, that's a strawman, this is just an alternative over the wire method (as is bluetooth file transfer)
Classic lore! :)
The joke at hand. [0] Here we go again.
Now let see if the founder(s) will reply here and to run it all back from start to IPO.
What would the founders do differently from dhouston this time?
[0] https://news.ycombinator.com/item?id=9224
Haha
> you can already build such a system yourself quite trivially by getting an FTP account, mounting it locally with curlftpfs, and then using SVN or CVS on the mounted filesystem...
I don't think the word "trivially" means what you think it means.
[Edit: I now realize the above is a verbatim quote from a naysayer after the Dropbox announcement]
Other comments have added context, but this is a Hacker News "copypasta" of sorts, from DropBox's first launch.
> I don't think the word "trivially" means what you think it means.
You cut off "for a linux user [in 2007]" and that's a very important part of the sentence.
[dead]
"former dropbox engineers" doesn't mean a whole lot -- it's a large company where tens of thousands of people have worked over the years. It's not like this is by the founders or anything.
Yea, "former [COMPANY] employee" could mean anything. I'm not sure it's really much of a flex. I'm a former Apple employee. Nobody gives a shit, and nobody should--That doesn't count for anything if I were to do a software startup. It wouldn't even bear mentioning in a press release.
Ok, we've taken the former dropbox engineers out of the title now.
So airdrop…
Not sure why this needs to be a service. Or why data needs to go through their servers.
It looks like this lets you share files with people using a different operating system than MacOS. Does AirDrop let you do that?
His point is about it being a service vs just a peer to peer app
So you run the app. The other person runs the app.
How do those 2 programs find out about each other?
Well, they need to use some server that will connect all users, allow you to find the other person in order to send them a file.
Even BitTorrent needs tracker servers to connect downloaders to uploaders.
Now, you could try to make it run without a company behind it, like BitTorrent.
I guess it's an exercise for the reader.
Someone still needs to run those connecting servers out of the goodness of their heart.
You would need to degrade the usability (the first thing a user would need to do is to configure the app with the address of at least one connecting server).
BitTorrent is anonymous but for sending files you need to connect to a person you know. If you enable contacting via the protocol, it'll be a phishing nightmare.
So you'll have to degrade usability some more and use a side channel (e.g. an email) to exchange identities with people you want to send file to (or receive files from).
By having a service and a company behind it, they are able to perfect the UX and run the necessary servers to implement that UX.
Universities, the Debian Foundation, Nvidia, the ccc, hell even Microsoft run magic-wormhole relays... the only thing new here is the subscription and the hipster gui client.
This is classic critique. Dropbox was criticized in the same way when it was announced years ago.
Magic Wormhole is cool (I just learned about it! Thanks! :)), but most people don't use nor care about the command line, and are probably afraid of it.
I am more in critique of the shameless subscription for services like this. US$25 /user/month for what? Certainly not the traffic.
I think that might be billed on the sender vs the receiver. So if you're a solo creative person sharing stuff with clients, this can be one way of doing that. Also potentially useful for small businesses without an enterprise file sharing solution.
It appears that the data is relayed as a fallback, if there's no p2p connection possible. This is a classic transfer model for mediated transfers.
AirDrop only works on local networks, not over the Internet.