(Yes, I know about the extension that hijacks your searches to redirect them to Kagi, but how is that an acceptable state of affairs?)
I wanted to use Safari at work, but this proved too big of a barrier. I can’t use the App Store at work, so no extensions. I was more willing to give up Safari than Kagi.
I’m requesting that iOS Safari allow users to set any search engine as their default, rather than limiting choices to a pre-approved list. Currently, Safari only offers Google, Bing, Yahoo, DuckDuckGo, and Ecosia, preventing users from choosing legitimate alternatives like Kagi.
While workarounds exist, they’re cumbersome and don’t provide a seamless experience. Please consider allowing users to add custom search engines as default options, similar to macOS Safari. This would enhance user choice without compromising security.
Thank you for considering this feedback.
My dream is some of boot-to-Ubuntu setup, or even running it simultaneously.
I am pretty that is exactly the point for Apple.
The extension is working really well for me. You could always try it out with the free version to see if it works for you as well.
[0] Not on the home screen, but I'll take what I can get.
And sure enough, Safari on macOS seems to not allow it at all (needs extensions).
Does Kagi plan to open-source Orion on Linux?
I get (though wouldn't necessarily agree with) keeping it closed while it's still in the works, but would like to know if the plan is to open source in the future or not.
Jeez, downvoted for asking about context? People, calm down.
For me, it's a big deal (although not a dealbreaker) for that reason. If I have the option of two different pieces of software, one being open source and the other proprietary, I'll choose the open source one every time unless there's something really exceptional about the proprietary one. But that's very rare.
I was just trying to think of any proprietary software I use outside of work (where I don't have a choice) or games. There must be at least one, but I can't think of what it is.
Personally, I think it would be incredible if you open sourced your search engine. But like someone else said more eloquently, software runs on our computers. And to me, open-source software is table stakes when there are viable alternatives.
But since I have your attention, I just want to add that I'm a huge fan of Kagi Search and it's well worth the money I spend for it. I love the work you guys are doing, and that love is the reason why I'm even thinking about using Orion. But they are two entirely different use cases.
There aren’t great open-source search engines, so I’m moving from one proprietary option to the next. But there are great, open-source browsers already, and I refuse to go backwards.
If a good, open-source search engine were available, I would leave Kagi for it.
Do I? I'm not going to post sensitive information into a search engine no matter who runs it.
My search history ain't worth much. What the contents of e.g. my bank website are is.
Yes, it's a big deal. I've lived in the non-free software world before and struggled to get out. I'm not going back.
Not necessarily, Kagi provides a feature[1] that anonymizes all your searches. I set it up and haven't thought about it since.
You’re still trusting them. Not to mention they could round them all up by IP or browser fingerprinting.
There is still some level of trust.
I happen to trust them enough for that; but it is still trust.
https://blog.kagi.com/kagi-privacy-pass#token-generation:~:t....
Where does it show that on the Kagi backend they couldn’t, theoretically, save the session key before performing the token response?
If you're not going to make an effort to understand how it works, don't make assertions about how it works. Ask your favorite LLM about the RFC if you have any further questions.
I really hope Kagi contributes back upstream to improve the situation, it’s needed.
Edit: looks like konqueror uses qt web engine which is chromium. The irony of the KDE browser abandoning WebKit while the GNOME browser still tries to use it is too much.
We also see limited upside from community contributions - the number of people who can meaningfully work on a WebKit browser is small (from our experience hiring), and most of them already work at Apple or Kagi. Meanwhile, managing an open source codebase of this size would add real strain to our small team.
The plan is however to open source when Orion is self-sufficient (business model of Orion is you are the customer and can pay for it - like we used to pay for browsers 20 years ago before advertisers started paying for our browsing), meaning it can sustain its own development independent of Kagi Search. I want to take the opportunity to thank all people who supported the Orion browser vision [1]. We're not there yet but recent 1.0 launch and expanding to Linux are steps in that direction. And on Jan 1st this year we began development of Orion for Windows (HN exclusive yay!).
I understand this is unsatisfying to people who want source access now. It's a tradeoff we've made deliberately, not something we're hiding behind.
> Kagi founder here. Orion isn't open source yet primarily because we're a 5-person team that spent 6+ years building this and created significant IP doing so,
But it's possible I haven't considered some detail where I might agree it's reasonable. Can you describe or offer any insight into the "significant IP" that you need to protect and defend? What threats from a larger company are you primarily concerned about?
We believe software and hardware creators have a right to choose their business model and let that model compete, as Kagi's is competing right here in this thread.
* Having worked at mega banks etc., they do look at these numbers to decide whether to invest in standards support or slap on a "Requires IE" button.
For a lot of people (even relatively geeky people), their computers end up being "an interface to use a browser". People use their browser to file their taxes, to write their documents, to manage their websites, to create websites, to look at porn, to pirate movies, to chat with their friends, to send/receive money to their bank, and a whole bunch of other things.
It would be hard to imagine a piece of software that is capable of knowing me more intimately than my primary web browser, and as Google has proven, this intimate knowledge is valuable. Companies pay boatloads of money for large quantities of personal information to target ads (and probably a bunch of other more disturbing things).
I genuinely don't think freediver is lying; I believe him when he says there's no telemetry data being sent and that it's not tracking me, but there's the sticking word: "believe". I have to trust him, which wouldn't necessarily be the case if it were FOSS.
Now, granted, I could always run Wireshark or something to ensure that there's no telemetry data being sent regularly, but that only protects you so much; for all I know, they could be taking steps to actively make it look like they're not sending data, or they could be batching up N days of data and sending it in batches so it is not as obvious that telemetry is sent.
Again, I genuinely don't think they're doing that, I believe them, but I do see peoples' points.
I understand your position, but a web browser is so important a software that it must be open.
I also think that you can still sell it even if it is open source.
Also, you might be able to secure funding from governments that want to move away from closed source solutions.
Anyway, still congratulation for v 1.0, and I hope it will go well.
The state of webkitgtk is a bit rough, as I’m sure you and your engineers have noticed. The other part of what open source means to people is that you contribute back to the open source code you used to build your business, lifting all boats in the process.
What people certainly do not want to see is Kagi pull an Apple: utilize FOSS to the extent it helps you but return nothing but “thanks everyone but we got ours”.
And when the market is going to be primarily technical people I don't think you can trust them/us with source-available either as hackers with a strong aversion to paying for software thinking themselves clever will make and distribute bootleg builds with the license checks removed. Then you'll have to spend your time finding and DMCAing them which will only make people mad. Best to avoid it entirely.
I appreciate you/Kagi actually thinking about building a sustainable business in contrast to companies that open source their core competency and then fail to make money later.
Source: happy paying customer and user of Orion.
Can you please elaborate what do you mean when you say this? This is something I do not understand. How licensing terms affect your codebase management beyond setting things up so the code is available to users?
Publishing something under a FLOSS license doesn’t mean anything except that you grant end-users certain rights (the four essential freedoms). The rest (like accepting patches or supporting external developers) is customary but by no means obligatory. You don’t have a capacity for it - don’t do it, easy. There are thousands of developers who do that - they just dump whatever they have under a nice license and that’s it.
Unless you’re saying your legal department doesn’t have capacity to handle licensing concerns, especially if you’re using or potentially using non-FLOSS third party components. That I can totally understand, it could be pretty gnarly.
Please don’t be mistaken: Free Software is a purely legal matter of what you allow users to do with your work - not some operating principles or way of organizing processes.
Note: All this said, I can understand that you may not want to grant some freedoms to the end users, particularly the freedom to redistribute copies, because this could affect your plans of selling the licenses. But that’d be a whole different story than codebase management concerns.
My read is their legal department isn’t fleshed out enough to defend the work when e.g. a tech giant steals it.
Orion will never reach "self-sufficiency" as long as you don't actually charge for Orion. Orion is completely free to use. I can donate to Orion+, but Orion+ offers no paid features; it's basically a Patreon. https://help.kagi.com/orion/orion-plus/orion-plus.html
(No major browser has ever sustained its own development independent of a search engine's funding, not even Netscape, which charged $40/seat in the 1990s, with a free "shareware" tier so generous that hardly anyone paid. Netscape was funded by advertising, especially from Yahoo search. Funding browser development entirely on donations to a commercial business would be completely unprecedented.)
What if, instead, you made Orion "source available" to paying customers, but not open source? You could merge PRs only from users who sign a CLA. (Users would file PRs out of charity, for the same reason they sign up for Orion+ today.)
EDIT: I found this in the docs:
> The alpha version of Orion for Linux is currently only available to [Orion+ supporters](../orion-plus/orion-plus.md) and can be downloaded from the [Billing Dashboard](https://kagi.com/settings/billing) under the Orion browser section.
So I still can't test this. Only for Orion+ supporters.
Regardless, great to see this come to Linux.
https://orionfeedback.org/d/2321-orion-for-windows-android-l...
It is interesting that they don't mention future Android support anywhere on the website, but it does seem like they are (or at least were) open to the idea someday.
We talk a lot about browser diversity, but on Linux and Windows, it is a lie. You have firefox (gecko) and fifty flavors of chromium. Webkit on Linux has essentially been relegated to embedded devices or the GNOME epiphany browser, which I'll admit while is a noble effort, lags a bit in the stability and power-user features department. Big reason for that is that it lacks the commercial backing to keep up with the modern web standards rat race.
Kagi bringing orion to Linux changes the calculus. It introduces a third commercially incentivized, consumer-grade engine to the platform. Even if you never use orion, you want this to succeed because it forces WebKitGTK upstream to get better, which benefits the entire open source ecosystem.
The sticking point like always will be media playback (read: DRM/widevine). That is the graveyard where Linux browsers go to die. If Kagi can legally and technically solve the widevine integration on a non-standard Linux webkit build, they win. If not, it will be a secondary browser.
Probably true in general. But for me, that's not a sticking point at all. I don't care if a browser supports media playback or not.
What I do care about is the ability to enable/disable embedded code execution (JS, at the very least) at a fairly granular level. Does Orion allow for that?
If users are unwilling to opt out of that abuse then I think its OK that their migration to Linux remains mildly inconvenient.
The time to thrash against DRM will be when you can't get hardware that boots kernels which were not approved by your government. What we're doing now is trying to prevent that time from coming at all.
Having to watch protected media on a different device is a small price to pay.
Speak for yourself. I avoid any DRMed content. If I pay for it, I'm getting unrestricted access to the files. Or I don't pay for it.
My take is: you're welcome to the party, but don't be surprised if someone shows you the door when you pull out the drugs. It ain't our type of fun.
I understand various reasons why people are pushing for the adoption of open source software, but it will be counterproductive if it brings the problems of the commercial software world with them.
I'm hopeful that some day Linux will have enough users where the media companies can't ignore them. Hopefully, that day is sooner than later.
It's pretty frustrating that peacock (and all xfinity streaming) doesn't work and you can't get 1080p or 4k on most other streaming platforms.
Correct me if I'm wrong but to stream 4K, studios require a hardware root of trust and a verified media path. They need a guarantee that the video frames are decrypted inside a trusted execution environment and sent directly to the display without the OS kernel or user space being able to read the raw buffer.
AFAIK Windows and macOS provide this pipeline at the OS level. OTOH, ChromeOS gets 1080p/4K not because it has massive market share but cause the hardware and boot chain are locked down by the almighty Google.
On desktop Linux, where you have root access and can modify the kernel or compositor to inspect memory, there is technically no way to guarantee that secure path to the studios' satisfaction. Am I right in this assumption?
Unless the DRM providers change their threat model, which sounds unlikely to me. Or distros start shipping signed and locked-down kernel modules that prevent the user from being root, which is again unacceptable to most (me included), we will likely be capped at 720p for some time now.
Yes. I tried using Chrome on Linux just to watch movies that I purchased on Youtube at HD/4K and watched as the stream was limited to 240P. IMHO regardless of what Google says in their ToS they have already broken the trust agreement by not providing what I paid for. Regardless of what the studios want, all this does is push me back towards piracy because once again the industry fails to understand that piracy is a accessibility problem, not a financial problem. If I pay for 4K then regardless of where I want to watch that movie it better be in 4K, that's what I paid for. Google hides behind their ToS to get around the fact that they sold me a product then failed to deliver.
> ChromeOS gets 1080p/4K not because it has massive market share but cause the hardware and boot chain are locked down by the almighty Google.
ChromeOS is based on Gentoo Linux underneath just very stripped down and Googlefied. It's the same BS that Bungee pulled with Destiny 2 and Linux. If you so much as dared to run Destiny 2 on Linux you would be banned. Stadia used Linux but because Google controlled the platform they allowed it to be played there.
These are the games they play to make other platforms that aren't MacOS/Windows appear like they are incapable but in reality it's just corporate greed and grift.
Oscilloscopes and signal analysers exist.
s/need/want/ but yes.
Does YouTube and Netflix work? That's the lion's share right there. A lot of users probably don't even care about the other streaming platforms. I'm probably being too optimistic, but I think the upcoming Steam machines will have a significant adoption of the linux desktop. Microsoft is certainly working 'round the clock to alienate their users.
The same is true for running a vanilla Chromium build on Windows, the big difference is the quality of content you can get on Windows can be higher than 720p in the mainstream browsers (as long as the rest of the system is compliant as well).
I agree it's a bit silly, but I think a lot of people don't really care about quality so long as they can watch it. I guess that'd also explain how Netflix gets away with such low bitrates for even the "high quality" versions of content.
- 15% PC
- 10% Smartphone
- 5% Tablets
- 70% TVs
In terms of viewing hours https://www.statista.com/chart/13191/netflix-usage-by-device.... So definitely most viewing on TV, but still something like 1/3 of households with TVs don't have a 4k TV at all (as of 2025) in the first place. Hard to definitively say more since Netflix & others don't seem to publish the numbers often.
I'd love to find out I'm wildly wrong though and have a bunch of people willing to push Netflix to have higher quality content... but so many people don't even seem to pay for the premium plan with 4k (anecdotally, Netflix doesn't seem to publish numbers on that) that I'm not holding my breath as I sit here with UHD Blu-Ray quality instead :D. It seems like most people just want something quick to turn on in the background than something to really sit down and bask in every detail of.
Right. The user I was replying to was asking about a browser that isn't either of those.
One correction to my message above: apparently Chrome on Windows is still 720p for Netflix, it was Edge that had 4k support. Or you can install the Netflix App on Windows too.
Most people will not use a browser that can't open youtube videos and they know and exploit this with extreme precision.
You're already paying the monthly fee to stream it, you're just streaming it in a more friendly way. Granted if you cancel the service, you should delete the content.
Many won't though and that's the problem but that problem is caused by the fact that you're being restricted in the first place.
Same situation on Mac where Apple's Fairplay DRM enables 4K playback in Safari, but Chrome and Firefox have the same limitations as on Windows.
Last time I tried to use Firefox on Windows as my daily driver, video playback was one of the biggest gaps that made me go back to Edge.
Maybe it's not widevine L1 but Firefox has the widevine plugin enabled on my Debian 13. I don't remember I had to do anything except downloading Firefox from Mozilla and installing it.
https://support.mozilla.org/en-US/kb/enable-drm
Apparently it's part of Brave too but it's disabled by default
https://support.brave.app/hc/en-us/articles/360023851591-How...
I expect it to be available on Chrome and I don't expect much from Epiphany.
> If not, it will be a secondary browser for documentation reading only.
I don't even have sound on my main desktop PC: the one I use the most. The one I do all my "life admin" stuff from (banking, real estate, etc.), all my work emails, all my coding. I think sound works but I haven't bothered to plug in speakers to check (since three years, when I assembled the PC).
That's a bit more than documentation reading.
There are work environments where even just a sound emitted by a PC is frowned upon.
People who aren't into media consumption are not just "reading documentation".
Don't forget about https://falkon.org. It's a browser I enjoy using. WebExtension support will be big if it lands in Orion though.
EDIT: apparently Orion is not open source. Not particularly interested in a closed source browser, TBH. In 2022 they said they plan to open source "when there is merit"[1], whatever that means. No merit yet, it seems.
[1] https://orionfeedback.org/d/3882-open-source-the-browser/2
On Firefox, you can disable DRM in about:config. Forks such as Librewolf and Tor Browser disable DRM by default.
That said, closed source is still a deal breaker for a browser for me, or I’d probably already be using Vivaldi.
What's the integration you're referring to here? This was something I was interested in, but as far as I can tell if I enable Kagi Privacy Pass it's enabled browser-wide, not just in private tabs.
WebKit development is mainly driven by Apple, which isn't great from the perspective of having a browser free of corporate influence, but I trust them more than Google.
It uses webkit, which is what Blink was forked from but likely has more in common with Safari at this point.
Homebrew also initially had mac only support, later Linux. But it always felt as if Linux was a second-class citizen. Is that also the case for Orion?
I’ve picked it up again as my daily driver as of the new year and haven’t had a single issue yet. It even blocks ads in YouTube now - only Brave did that previously.
For me - Brave was the best browser product. It’s ad blocking is truly phenomenal and nearly every site “just works”. But I don’t love the ethics of Brave and certainly not its founder. So I am extremely excited to have Orion take over that niche of the browser space that I most care about.
What would you say has changed over the past few months? I just felt like Kagi wasn't prioritizing Orion development enough, being busy with their main Kagi subscription and all.
In the meantime I'll give Orion a try as soon as they introduce cross-device sync.
I really want to switch, but no 1P support makes it really hard, unfortunately.
"Can" is doing a lot of work here. A browser's whole purpose is to be online, after all. If they were trying to collect information about you, they really don't need the installer to be the thing that does it. It would be an impressive reversal of their whole premise as a business if their browser's installer was the piece that was violating your privacy and not, you know, their whole service (that you have to be signed in to).
Not its whole purpose. I use browsers offline fairly often.
Offline installers (for any piece of software) are important to me because they allow me to keep a backup of the installer and won't restrict me when I don't have internet access. Keeping a backup is important because it lets me install older versions of the software when needed.
They do, if they are being duplicitous about their intent to not harvest user data.
If their browser was a data harvester from the get go, no one who is aware, and worried, about surveillance capitalism would have bothered to use it. And note that they had no problem in offering offline installers in the beginning. Now, once their base has grown, if they have a malicious intent (now or in the future), they can use the online installers to gather our personal data surreptitiously - for example, by profiling our hardware and (if you already had Orion installed) our settings, our bookmarks, our browser history etc. and use that commercially. It also allows them to install unwanted software on our computer in the future (I don't know if you are old enough to know - look up the browser toolbars era).
If their intent to respect a user's privacy is honest, offering an offline installer shouldn't be a big deal. (As far as I am aware, apart from Apple Safari, they are now the only browser that don't offer an offline installer).
The weirdest choice at the moment is by default Kagi sends a referrer when you visit a search result. There's currently ~65.000 Kagi subscribers worldwide, so just that lone data point completely destroys any anti-fingerprinting you're doing. And probably these subscribers are divided among time regions, so not all are active at the same time.
Even if you are on a VPN and visit site #1, then site #2, you are already cross-site trackable because it is very unlikely you are on the same VPN vendor (and endpoint) as the other subscribers. If you add in more data points like browser, OS, screen size and the like it becomes even more grim.
They have the referrer enabled because it helps make admins aware I guess.
You can turn it off (Settings > Privacy > Hide Kagi referrer), but defaults matter.
What are your thoughts on upgrading gnome 48 to 49 as a dependency?
I must +1 that no matter the platform (this criticism is not limited to Linux), the open source option is almost always my choice, especially for something as important as a browser. This is a big reason I don’t use Orion today, even though I have big issues with the other available browsers.
But what I really miss is a self-hosted sync server. I don't want to use a browser without sync, but I also don't want to trust this data with any 3rd party other than myself.
It's one of the main reasons I'm using Firefox, since that is the the only browser that even vaguely supports this - albeit not well.
I'm not interested in using a proprietary browser, and hope for a release under a free license at some point. But a free WebKit-based browser with Web Extensions could have interesting properties regarding battery life on mobile GNU workstations.
It probably matters less to Linux users who do the minimal tiling WM thing, but as someone more drawn to traditional floating DEs it's always bothered me how alien the browsers one might actually want to use feel running in a GTK or Qt DE. Themes can help reduce the gap, but it never disappears — that last 10-20% always remains as an unavoidable side effect of how the big browsers are built, with it being particularly pronounced with Firefox and derivatives.
Of course a GTK based browser like Orion isn't going to feel the best under a Qt desktop like KDE, but GTK themed to match Plasma is a good deal closer than the bespoke UI found in e.g. Firefox or Chrome.
Funny enough, Epiphany used to be a more power user oriented browser, and it used to be powered by none other than Gecko. Unfortunately, Mozilla killed embedding in Gecko and that (along with related projects like Camino and K-Meleon) came to a screeching halt and Gecko became hard coupled with Mozilla's UI decisions.
This is a huge disappointment.
1. It's a lot of work to maintain an open source project accepting community contributions. Absolutely true, but that's not what's being asked for. Providing a tarball under an open source license doesn't add any significant work. 2. No one has asked for the Kagi backends to be open sourced, so why is the browser different? Obviously because I run the browser on my machine. Your backend runs on your machine. 3. We need to protect our IP. Then release it under a copyleft license. Or if you absolutely must, release your proprietary bit under a non-open source license. 4. You don't need the source because we send 0 telemetry, which you can verify using a network proxy. That's hardly the only thing to be worried about with a binary blob. Even if you kept the code completely closed source, by just releasing a tarball with the source under a proprietary license, I can build my own binary from source and eliminate this threat.
An old mentor once said to me that a contract is just the start of a conversation. If you sign a contract, the other party violates it, and your business goes under... you may be able to get some compensation through courts, but also your business is gone. And getting that compensation and proving that the contract was violated and how much you are entitled to costs time and money.
Releasing something at all, even under a restrictive license, means nothing if you have no intention (or capability) of enforcing that license. Look at how often companies take GPL code, modify it, and then never publish their modifications... and then people have to sue to get things resolved.
So "We aren't ready to commit the legal resources to fighting and defending the licenses" makes a LOT of sense. IP protection is not just a matter of signing a piece of paper saying people can't do a thing, you have to actually prevent them from doing the thing.
No affiliation with Kagi, but I think you're dreaming if this actually would make a difference.
How many times has GPL'd software successfully been argued in court? Maybe four or five? Considering how many millions of software packages exist and how hard it would be to prove enough to bring a lawsuit/discovery request, I would be extremely surprised if there aren't thousands of GPL violations out in the wild that never go to court. I remember the source code to Spongebob Squarepants Supersponge violated the GPL [1], and that wasn't discovered for decades.
I am mostly ok with FOSS, and I don't love the idea of using a fully proprietary browser either, and I am probably not going to use Orion on Linux, but I don't think it's inherently wrong for them to want to keep any secret sauce close to their chest.
I tried Orion in Dec 2025 and it crashed more frequently than I could handle. It and/or VPNs I was using worked poorly with websites that I want to keep a page open on for days at a time.
I don't know what they do, but it caused weird graphics glitches and kernel panics simply from running in the background.
(They don't seem to be going for the "serious, clueful privacy and Internet freedom" demographic, or they wouldn't be using Discord.)
sergiotapia•12h ago