I have to imagine assistant (AI usage) will be next, and swap to pay as you go. And search won't get cheaper because I imagine they've made the mistake of treating the unused AI credit as a subsidy for other services
Our $10 search sub price is solely based off of the cost of operating search and has remained unchanged for years (est. 2023), before we offered any LLM-based features as part of our plans.
We added AI features to the existing plans at no extra charge. The only thing we preempted was our Ultimate plan (which we initially offered with literally zero extra perks!), anticipating that LLMs would become more capable and popular and require a higher margin to operate at scale.
The above two only disabled free access, due to costs. They mention that Kagi Translate would become a subscription service (so I pessmistic'ly assumed it would no longer be part of my existing subscription).
The 2026-29-10 announcement includes "As we roll out our unified Fair Use system, usage across Kagi Translate and Kagi Assistant will count toward a shared allowance.".
> Unions are very much an in-group vs out-group phenomenon (and in many cases, the benefits are specifically to the more senior union members vis a vis the less senior
In the US. In the UK all the major unions I know don't require membership and fight for all employees in the industries they represent. E.g. as a software developer on a research project at a University I got all the negotiated benefits of the academic Union.
In Italy too. We have national contracts that are negotiated by unions for the whole indistry and are valid for all workers of the same industry, not only members.
>I would much rather be able to have things like personalised searches (which are identifying).
This wouldn't affect your accounts usability or features it would just delink your payment from your account. Same as paying for Mullvad with a giftcard bought on Amazon.
Privacy Pass would decouple each search from the account. I think it’s okay to have as an opt-in thing. But private payments should be the default IMO.
I see. I mispoke with "privacy pass" although it's similar. I meant like a one time break between your payment and account for regular accounts underneath the option to use privacy pass.
I think we’re broadly in agreement: the more private options we have (with different levels of privacy), the better.
It’s possible, in theory, to use Privacy Pass to allow e.g. using credit card to pay for the whole subscription without associating the account with the payment. (E.g. one payment gives you 1 “pass”, and it can be used to extend your account validity for one month.) In practice, it’ll probably be fairly easy to correlate a payment with an account extension, as those would happen almost simultaneously, so a couple more levels are needed here (e.g. doing this through resellers, and maybe breaking up the subscription payment into, say, 30 “day passes” which are used in random order to make it harder to track).
Posteo’s “one-time code system” is a fair bit weaker than Privacy Pass btw: when they get a payment, they can still associate it with your account (so if you’ve paid using, say, bank transfer, they will know your name, and you’ll have to trust them not to store the link). But it can be used together with Privacy Pass :-)
This shows 10 credits per 5.6 Sol chat message, but the price for 100 credits I see on my (consumer) account is just $4. Are business credits vastly more expensive?
Everyone gets $125 / month budget (compared to the $5k/month people get for coding LLMS) - it's pretty rare for anyone to get even close - maybe some of the administrative assistant who do nothing but type into CoWork all day go over - I think I use about $30-$50/month.
Enterprise pricing: 10 credits per message if not "Instant". Our monthly charge for 30k credits makes that around $2.
I might be out by a factor of 2/3 (automatically applied promo deals/credits make it very hard to be sure), but we get 3k messages per month for a seat count of ~60, and finance are concerned about cost.
$2 is probably a good deal for the Extra High 8 minutes of agentic searching/coding/etc when it one-shots a whole project.
$2 is a terrible deal for a 5 second instant retrieval.
I just have an always on VPN so can tunnel back to my home assistant server; which can handle todos/reminders/calendar events. Quick messages to contacts I haven't considered.
I already use a VPN (Proton) so I can't VPN back to my home (i.e. Tailscale). A very unfortunate limitation of Android.
RethinkDNS has the potential to solve all my problems since it can do per-app proxies, however I've found that it doesn't work well with Proton (or possibly Wireguard as a whole) with frequent proxy disconnects, making it a very painful experience. Which is a shame because in some cases I value DNS filtering more than VPN.
The latest version is v055z [0], which handles WireGuard re-connectivity better across a range of providers. One Proton specific quirk we've hit was: Since Rethink lets you connect multiple WireGuards at once, their free tier especially may start rejecting "WireGuard handshakes" from your client, if you do so.
Some tips:
- If you don't mind extra battery use, setup "Persistent KeepAlive" for your all your WireGuard Peers. Any value between 30s to 100s is decent enough.
- Alternatively, if you're on version v055v+, then consider turning on Configure -> Network -> Smart Persistent Keepalive which shouldn't use up as much battery but yet attempt to keep a connection open whenever "user presence" is expected by Rethink.
- Turn on Configure -> DNS -> Split DNS only if you're Android 12+ which has comparatively better APIs to be able to implement per-app split-tunnel DNS across multiple WireGuards.
- Tap on the "Refresh" icon in Configure -> Proxy -> Setup WireGuard to instruct Rethink to attempt re-connections to all dormant Peers.
- Look for internal state in About -> General (scroll to the bottom to the section that goes "WireGuard") and see if you spot any mismatch there with respect to MTU, for example. And if so, consider opening a report on GitHub.
- Look for About -> App Logs and see if you spot any errors (or for warning, info, or debug, you'll have to change the log level when prompted) related to WireGuard. You can use the search bar to filter for "wg", which should keep the output limited to WireGuard stuff.
- Change Configure -> Network -> Choose IP version to "Auto" if you happen to go between IPv4-only / IPv6-only / dual-stack networks, or set it to "IPv4 & IPv6", if you're always on dual-stack networks.
Thanks for the tips. I currently have 0.5.5y - I'll update and try these out and see if my experience improves. I also have a paid Proton plan so they shouldn't be rejecting my connections but it's definitely still possible since I think their plans are for "up to 5 devices" or something like that, and I do have a few connections at all times.
Works for me in my dedicated Android Auto user profile in GrapheneOS. Proton VPN with 'killswitch' turned on, and I've been able to use Android Auto. Though, I don't think I gave Android Auto network/internet permissions, so it shouldn't know if I'm on a VPN or not. Android Auto doesn't need internet to function (though the apps you run in it might, i.e. Google Maps).
The ring can record for 12-15 hours before it's empty. I'd guess it's incredibly rare that people throw away their phone before reaching 15h * 20 = 12.5 days of active screen time.
I personally don’t find it that useful for most tasks, but if say, you get paid $50/hr for your work and it saves you more than 4 hours of work in a month, there you go.
Obviously this assumes that you can find 4+ extra hours of $50/hr work every month, or you can work 4 hours less. Neither of these assumptions is correct for people who work for a fixed salary.
I think this is the rub the enterprise will be forced to grapple with. Not everyone is going to get $200 worth of value for the organization. In fact since it's not a restricted tool some will waste time and company resources using it. Undoubtedly some will get the value out of it, but it's very likely, that these are the same people providing more than what they're paid already. Nothing has changed other than, potentially, time savings and (hopefully) output improvement. Neither of those are any sort of guarantee though, either. Subjective systems are hard to show value, especially in the long term.
If you are not working paycheck to paycheck, but still need to commit that time to work, you may prefer $200 worth of free time that you otherwise could not have - that is my general point of view there.
Here most of my colleagues have +200 dollar rates. It's really a no brainer. But sure, in south America or some Asian countries maybe it is. But still most devs need it anyway. Also in the poor regions.
$200/h is on the extreme end and I would argue most people here aren't anywhere close to that.
The median hourly wage in the US is $28/h, this equates to nearly 7.5 hours. A full day of work a month for the average person to use Claude with reasonable limits.
Yes, the people on $28/h may not be the software development types, so their income might not be as high, but these are the people who would probably be vibe coding the most since they aren't day to day programmers!
In Sweden $200 is ~5% of average programmer monthly income after tax. $200/h rate is not a representative salary for SEs in South America, Asian countries nor Europe.
If you're running a business I agree it's a no-brainer, but the context here is for personal projects.
reply