[HN Gopher] Google AI Ultra
___________________________________________________________________
Google AI Ultra
Author : mfiguiere
Score : 201 points
Date : 2025-05-20 18:20 UTC (4 hours ago)
(HTM) web link (blog.google)
(TXT) w3m dump (blog.google)
| codydkdc wrote:
| I really want Google to launch a Claude Code/OpenAI Codex CLI
| alternative. also if they included a small VM in one of these
| subscriptions I'd seriously consider it!
| boole1854 wrote:
| They are working on it: https://jules.google/
| unshavedyak wrote:
| I got the feeling Jules was targeted at Web (ala Github) PR
| workflows. Is it not?
|
| The Claude Code UX is nice imo, but i didn't get the
| impression Jules is that.
| kridsdale1 wrote:
| At Google, our PR flow and editing is all done in web based
| tools. Except for the nerds who like vi.
| codydkdc wrote:
| people don't use local editors? it's weird to lock people
| into workflows like that
| johnisgood wrote:
| Damn... you guys don't use proper text editors?
| incognito124 wrote:
| https://firebase.studio/
| johnmlussier wrote:
| I'm not interested in about 75% of this offering. Really wish
| they had pieced this out or done a credit system for it all. I
| want higher coding and LLM limits, Deep Think, and would like to
| try agent stuff, but don't care at all about image or video
| generation, Notebook limits, YouTube Premium, or more storage.
|
| $250 is a the highest cost AI sub now. Not loving this direction.
| camillomiller wrote:
| What other direction would you expect to be possible? Even with
| this rates most AI companies are still bleeding money.
| piskov wrote:
| If anyone it's Google that would be very-very inference
| efficient (given their custom TPUs and what have you).
|
| However if all this power is wasted on video generation, then
| even them probably will choke.
|
| Then again, I guess your average Joe/Jane will looove to
| generate some 10 seconds of you daily whatsapp stuff to
| share.
| karmakurtisaani wrote:
| I wonder how long this free access to LLMs can continue. It's
| like early days of Facebook, before the ads and political
| interference started to happen. The question is, when will we
| see the enshittification of LLMs?
| camillomiller wrote:
| Exactly. When Masayoshi Son's money will run out I guess.
| Workaccount2 wrote:
| I'm assuming their will be an ala carte API offer too.
| causal wrote:
| Yeah it says "designed for coding" but it's missing the one
| thing programmers need which is just higher Gemini API usage
| limits.
| _heimdall wrote:
| Building and running these models is expensive, there's no way
| around that in the near future.
|
| LLM companies have just been eating the cost in hopes that
| people find them useful enough while drastically subsidized
| that they stay on the hook when prices actually cover the
| expense.
| linsomniac wrote:
| At $125/mo for 3 months I'm tempted to try it, but I don't
| understand how it would interfere with my existing youtube/2TB
| storage family plan, and that's a big barrier for me.
| paxys wrote:
| What's wrong with just using the API for that?
| logicchains wrote:
| Deep Think isn't available on the API.
| paxys wrote:
| Deep Think is _only_ available on the API. It 's just
| restricted to "trusted testers" right now before a wide
| launch.
| lallysingh wrote:
| $250 a month. Oh my.
| akomtu wrote:
| "Rent a brain for only $250/mo"
| SoftTalker wrote:
| About an hour of a senior developer's fully-loaded cost.
| paxys wrote:
| And less than an hour of an external consultant's time.
| piskov wrote:
| Given the lack of comments after an hour passed, we have a strong
| case of maybe five Google AI Ultra subscribers worldwide.
|
| I, personally, try to stay as far as possible from google: Kagi
| for search, Brave for browsing (Firefox previously), Pro on
| OpenAI, etc.
|
| We'll see how fair OpenAI will be with tracking and what have you
| (given "off" for improve for everyone), but Google? Nah.
| rohansood15 wrote:
| "I think there is a world market for maybe five computers." --
| Thomas Watson, chairman of IBM, 1943.
| sunaookami wrote:
| This is an urban legend btw, Thomas Watson never said that.
| __natty__ wrote:
| > YouTube Premium: An individual YouTube Premium plan lets you
| watch YouTube and listen to YouTube Music ad-free, offline and in
| the background.
|
| It seems weird to me, they included entertainment service in
| ,,work" related plan.
| jihadjihad wrote:
| They don't even have the decency to make it a family plan,
| either.
| dewey wrote:
| It's not though, it's just the highest tier of the regular
| "Google One" account that also has Google Photos etc. included.
| kumarm wrote:
| So everyone who want Youtube Premium can explain to their boss
| why they need Gemini AI Ultra for work?
| Keyframe wrote:
| Hmm, interesting. There's basically no information what makes
| Ultra worth that much money in concrete terms except "more
| quota". One interesting tidbid I've noticed is that it seems
| Google One (or what is it called now) also carries sub for
| youtube. So far, I'm still on "old" Google One for my family and
| myself storage and have a separate youtube subscription for the
| same. I still haven't seen a clear upgrade path, or even a
| discount based on how much I have left from the old subscription,
| if I ever choose to do so (why?).
|
| edit: also google AI Ultra links leads to AI Pro and there's no
| Ultra to choose from. GG Google, as always with their "launches".
| flakiness wrote:
| I believe Imagen 4 and Veo 3 (the newest image/video models)
| and the "deep think" variant are for Ultra only. (Is it worth
| it? It's a different question.)
| skybrian wrote:
| I just tried it and Whisk seems to be using Imagen 4 and and
| Veo 2 when used without a subscription.
| ComplexSystems wrote:
| The problem with all of these is that SOTA models keep changing.
| I thought about getting OpenAI's Pro subscription, and then
| Gemini flew ahead and was free. If I get this then sooner or
| later OpenAI or Anthropic will be back on top.
| SirensOfTitan wrote:
| This is even the case with Gemini:
|
| The Gemini 2.5 Pro 05/06 release by Google's own reported
| benchmarks was worse in 10/12 cases than the 3/25 version.
| Google re routed all traffic for the 3/25 checkpoint to the
| 05/06 version in the API.
|
| I'm also unsure who needs all of these expanded quotas because
| the old Gemini subscription had higher quotas than I could ever
| anticipate using.
| magicalist wrote:
| > _I'm also unsure who needs all of these expanded quotas
| because the old Gemini subscription had higher quotas than I
| could ever anticipate using._
|
| "Google AI Ultra" is a consumer offering though, there's no
| API to have quotas for?
| airstrike wrote:
| This 100%. Unless you are building a product around the latest
| models and absolutely must squeeze the latest available oomph,
| it's more advantageous to just wait a little bit.
| pc86 wrote:
| I am willing to pay for up to 2 models at a time but I am
| constantly swapping subscriptions around. I think I'd started
| and cancelled GPT and Claude subscriptions at least 3-4 times
| each.
| xnx wrote:
| > If I get this then sooner or later OpenAI or Anthropic will
| be back on top.
|
| The Gemini subscription is monthly, so not too much lock-in if
| you want to change later.
| Ancapistani wrote:
| I wonder if there's an opportunity here to abstract away these
| subscription costs and offer a consistent interface and
| experience?
|
| For example - what if someone were to start a company around a
| fork of LiteLLM? https://litellm.ai/
|
| LiteLLM, out of the box, lets you create a number of virtual
| API keys. Each key can be assigned to a user or a team, and can
| be granted access to one or more models (and their associated
| keys). Models are configured globally, but can have an
| arbitrary number of "real" and "virtual" keys.
|
| Then you could sell access to a host of primary providers -
| OpenAI, Google, Anthropic, Groq, Grok, etc. - through a single
| API endpoint and key. Users could switch between them by
| changing a line in a config file or choosing a model from a
| dropdown, depending on their interface.
|
| Assuming you're able to build a reasonable userbase, presumably
| you could then contract directly with providers for wholesale
| API usage. Pricing would be tricky, as part of your value prop
| would be abstracting away marginal costs, but I strongly
| suspect that very few people are actually consuming the full
| API quotas on these $200+ plans. Those that are are likely to
| be working directly with the providers to reduce both cost and
| latency, too.
|
| The other value you could offer is consistency. Your
| engineering team's core mission would be providing a consistent
| wrapper for all of these models - translating between OpenAI-
| compatible, Llama-style, and Claude-style APIs on the fly.
|
| Is there already a company doing this? If not, do you think
| this is a good or bad idea?
| planetpluta wrote:
| I think the biggest hurdle would be complying with the TOS.
| Imagine that OpenAI etc would not be a fan of sharing quotas
| across individuals in this way
| Ancapistani wrote:
| How does it differ from pretty much every SaaS app that's
| using OpenAI today?
| wild_egg wrote:
| Isn't that https://openrouter.ai? Or do you have something
| different in mind?
| Ancapistani wrote:
| I haven't seen this, but it looks like it solves at least
| half of what I was thinking.
|
| I'll investigate. Thanks!
| mrnzc wrote:
| I think what Langdock (YC-backed, https://www.langdock.com)
| offers might be matching to your proposal?!
| Ancapistani wrote:
| Looks like this is at least the unified provider. I'll dig
| in - thanks :)
| UncleOxidant wrote:
| You can just surf between Gemini, DeepSeek, Qwen, etc. using
| them for free. I can't see paying for any AI subscription at
| this point as the free models out there are quite good and are
| updated every few months (at least).
| devjab wrote:
| I wonder why anyone would pay these days, unless its using
| features outside of the chatbot. Between Claude, ChatGPT,
| Mistral, Gemini, Perplexity, Grok, Deepseek and son on, how do
| you ever really run out of free "wannabe pro"?
| smeeth wrote:
| I hate hate hate putting Deep Think behind a paywall. It's an
| ease-of-use tax. I fully expect to be able to get it over API
| through Poe or similar for way cheaper.
|
| Just have usage limit tiers!
| lvl155 wrote:
| Google is a bit tone deaf with these offerings. Are they
| interested in competing?
| vFunct wrote:
| Does this include Gemini 2.5 Pro API access? What are the API
| limits?
|
| I blew through my $40 monthly fee in Github Copilot Pro+ in a few
| hours. =^/
| Workaccount2 wrote:
| I suspect Google sees the writing on the wall, and needs to move
| to a more subscription based business model. I don't think the ad
| model of the internet is dead, but I also don't think it was
| particularly successful. People block ads rather than forgo those
| services, ads conditioned people to think everything on the
| internet is free, and the actual monetization of ad views makes
| people pretty uncomfortable.
|
| So here we are, with Google now wading into the waters of
| subscriptions. It's a good sign for those who are worried about
| AI manipulating them to buy things, and a bad sign for those who
| like the ad model.
|
| Is the future going to be everyone has an AI plan, just like a
| phone plan, or internet plan, that they shell out $30-$300/mo to
| use?
|
| I honestly would greatly prefer it if it meant privacy, but many
| people seem to greatly prefer the ad-model or ad-subsidized
| model.
|
| ETA: Subscription with ads is ad-subsidized. You pay less but
| watch more ads.
| aceazzameen wrote:
| It will eventually be subscriptions PLUS ads combined.
| continuational wrote:
| YouTube is probably the most expensive streaming app, and
| there are still ads (sponsored sections) in nearly every
| video.
| tintor wrote:
| Sponsored sections are baked into video and very easy to
| skip.
|
| Unlike platform ads which disable video control while the
| ad is playing.
| jeffbee wrote:
| Cannot find any factual basis for the claim that YTP is the
| most expensive streaming app. Is this the case in some non-
| US market?
| jerjerjer wrote:
| > ads (sponsored sections) in nearly every video.
|
| SponsorBlock for YouTube resolves the issue.
| conductr wrote:
| No idea if this is right but based on AI summary google
| results, annual operating costs for Youtube is ~$3B-5B and
| Netflix is ~$25B-30B. While YT probably spends most on
| CDN/bandwidth, they have a mostly free content cost which
| is by far Netflix's largest expense
| iamdelirium wrote:
| By that metric, every streaming platform has ads since they
| serve movies with product placement.
| ninininino wrote:
| They already do subs for YouTube w/o ads and for storage (email
| attachments + Google Photos + Google Drive), for Stadia while
| it was around.
| jeffbee wrote:
| This doesn't seem like new territory ("wading in"). This is an
| extension of the existing Google One plans to reach people with
| extreme demands.
| Etheryte wrote:
| I think this is a bit too rose tinted glasses. Being a paying
| customer doesn't necessarily mean you won't get ads, look at
| Netflix for a start. Their cheapest paid tier still gets ads.
| The subscription model will be an addition to the ad revenue,
| not a replacement.
| ljm wrote:
| It _should_ mean that though.
|
| Ads are well and truly the cancer on the service industry.
|
| It's an outright abuse to force ads and then make you pay for
| the bandwidth of those ads on your own plan to render them.
| pc86 wrote:
| Anyone can say A _should_ mean B, that doesn 't mean it's
| obviously true.
|
| Very few services still commercially viable today actually
| force ads - meaning there is no paid tier available that
| removes them entirely.
|
| I don't particularly like ads but this idea that _any_
| advertisement at _any_ point for _any_ good or service is
| by definition a cancer is a fringe idea, and a pretty silly
| one at that.
| myko wrote:
| Google used to let you pay a flat rate to avoid (most) of
| their ads. It was nice. This program was, of course,
| canceled.
| MichaelZuo wrote:
| How does this relate to the parent?
|
| I don't there was a claim that nobody would ever offer a
| partially subscription partially ad funded service.
| skarz wrote:
| I wonder how viable the referral link/referrer code method is?
| Based on my own YouTube viewing habits it seems like a lot of
| prominent channels have gone that route. Seems like it could
| work for the web overall. Ads would no longer have to target
| via cookies or browsing history because you could just serve
| links or offer codes related to your site's content.
| abtinf wrote:
| > I also don't think [the ad model of google] was particularly
| successful.
|
| Only on HN.
| narrator wrote:
| Yeah, only a few trillion in revenue over the last decades
| including Facebook and others. Not particularly successful.
| jonluca wrote:
| Actually hilarious, the distribution of comments on HN is
| truly bimodal
| Workaccount2 wrote:
| I mean that in that I don't think it lived up to what Google
| envisioned. People have extremely hostile views towards ads
| but have a full expectation that everything is just an
| account creation away, if not outright just given away.
|
| 30% of people who use Google don't view their ads. It's hard
| to call a business where 30% of people don't pay successful.
| The news agencies picked up on this years ago, and now it's
| all paywalls.
|
| This doesn't even get into the downstream effects of needing
| to coax people into spending more time on the platform in
| order to view more ads.
| tekla wrote:
| Maybe if you ignore objective reality.
|
| Google ads revenue AND income has consistently risen
| basically forever. Its ~75% of Alphabets total revenue and
| corresponds to over ~%50% of all Ad revenue in the world.
| netsharc wrote:
| Heh, although I'm a cheapskate, the ad-based world is a fucked
| up one. We now have an attention-economy, trying to keep you
| hooked on the content so "the platform" can serve you ads and
| earn money off you. And they do that by serving content that
| engages you, and apparently it's content that stirs up a lot of
| emotions.
|
| "Worried about refugees? Here's some videos about refugees
| being terrible". Replace "refugee" with "people celebrating
| Genocide", etc, etc...
| add-sub-mul-div wrote:
| > Is the future going to be everyone has an AI plan, just like
| a phone plan, or internet plan, that they shell out $30-$300/mo
| to use?
|
| Not the people who haven't been trained to require the crutch.
| kleiba wrote:
| These prices are nuts, in my opinion. It basically means that
| only companies can afford access to the latest offerings - this
| used to be the case for specialist software in the past (e.g., in
| the medical sector), but AI has the potential to be useful for
| anyone.
|
| Not a good development.
| esafak wrote:
| And I think it is a good thing. If there are buyers, it means
| they are getting that much value out of it. That there is a
| market for it. Competition will bring prices down.
| mschuster91 wrote:
| > Competition will do its thing and bring prices down.
|
| It won't. For now the AI "market" is artificially distorted
| by billionaires and trillion-dollar companies dumping insane
| amount of cash into NVDA, but when the money spigot dries out
| (which it inevitably will) prices are going to skyrocket and
| stay there for a loooong time.
| esafak wrote:
| How will prices skyrocket when there is a flood of open
| models? Or are you talking about GPU prices? They're
| already high.
| jsheard wrote:
| Who do you think is paying to train those open models?
| The notable ones are all released by VC-funded startups
| or gigacorps which are losing staggering amounts of money
| to make each new release happen. If nobody is making a
| profit from closed models then what hope do the companies
| releasing open models have when the money spigot runs
| dry?
|
| The open models which have already been released can't be
| taken back now of course, but it would be foolish to
| assume that SOTA freebies will keep coming forever.
| conductr wrote:
| It won't be the end of the world if the 'progress' were
| to slow down a little, I have trouble keeping up with
| what's available as it is - much less tinkering with it
| all
| delusional wrote:
| It will because "keeping up" is the sleight of hand. By
| constantly tweaking the model you don't ever notice
| anything it's consistently wrong about. If they "slowed
| progress" you'd notice.
|
| Current AI is Fast Fashion for computer people.
| johnisgood wrote:
| I do not think I will ever be able to afford hardware
| that is capable of running local LLMs. :(
|
| What I can afford right now is literally the ~20 EUR /
| month claude.ai pro subscription, and it works quite well
| for me.
| mschuster91 wrote:
| > How will prices skyrocket when there is a flood of open
| models?
|
| Easy: once the money spigot runs out and/or a proprietary
| model has a quality/featureset that other open-weight
| models can't match, it's game over. The open-weight
| models cost probably dozens of millions of dollars to
| train, this is not sustainable.
|
| And that's just training cost - inference costs are also
| massively subsidized by the money spigot, so the price
| for end users will go up from that alone as well.
| msikora wrote:
| ChatGPT is insanely subsidized. The $20/month sub is such a
| great value. Just the image gen is about $0.25 a pop
| through the API. That's 80 image generations for $20.
| jeffbee wrote:
| > used to be the case for specialist software
|
| I think that's a great example of how a competitive market
| drives these costs to zero. When solid modeling software was
| new Pro/ENGINEER cost ~$100k/year. Today the much more capable
| PTC Creo costs $3-$30k depending on the features you want and
| SOLIDWORKS has full features down to $220/month or $10/month
| for non-professionals.
| gigaflop wrote:
| Off-topic, but I work 'around' PTC software, and am surprised
| to see them mentioned. Got much knowledge in the area?
|
| On-topic, yeah. PTC sells "Please Call Us" software that, in
| Windchill's example, is big and chunky enough to where people
| keep service contracts in place for the stuff. But, the cost
| is justifiable to companies when the Windchill software can
| "Just Do PLM", and make their job of designing _real,
| physical products_ so much more effective, relative to not
| having PLM.
| Aurornis wrote:
| > It basically means that only companies can afford access to
| the latest offerings
|
| The $20/month plan provides similar access. They hint that in
| the future the most intense reasoning models will be in the
| Ultra plan (at least at first). Paying more for the most
| intense models shouldn't be surprising.
|
| There's plenty of affordable LLM access out there.
| Calwestjobs wrote:
| Magics !
|
| I do not know what that hate about 250 $ is, just flow is worth
| more.
| leoh wrote:
| I would agree, were I to use flow frequently; but I would guess
| it's the most operationally expensive API for Google and they
| may be subsidizing it (and profit in general) via users that
| don't use it (ie software developers).
| AJRF wrote:
| Would love to know how many people end up on this plan.
|
| If i had to guess, looking at the features I would have guessed
| 80 bucks. Absurdly high, but lots of little doodads and
| prototypes would make the price understandable at that price.
|
| 250?!
|
| I actually find that price worrying because it points to a degree
| of unsustainability in the economics of the products weve gotten
| used to.
| jeffbee wrote:
| The long-standing Google One plan with 30TB of storage was
| already $150/mo, so your estimate was a bit low.
| blagie wrote:
| I'm holding out for "Ultra Max Pro."
|
| (Comment is on the horrible naming; good naming schemes plan
| ahead for next month's offerings)
| kylehotchkiss wrote:
| Somehow, someway you're gonna need a Dell or Apple Ultra
| Edition to use it.
| ivape wrote:
| Is this the only price to get Google Flow at? Any alternatives?
| That seemms like the killer app here.
| adverbly wrote:
| Price point here is a bit too high... They have bundled so many
| things together into this that the sticker shock on the price is
| too much.
|
| I get what they're trying to do but if they were serious about
| this they would include some other small subscriptions as well...
| I should get some number of free movies on YouTube per month, I
| should be able to cancel a bunch of my other subscriptions... I
| should get free data with this or a free phone or something... I
| could see some value if I could actually just have one
| subscription but I'm not going to spend $250 a month on just
| another subscription to add to the pile...
| ehsankia wrote:
| They put anything that makes sense. I don't know if including
| random movies makes sense.
|
| They got Youtube Premium which is like 15$. 30TB of storage, a
| bit excessive and no equivalent but 20TB is around 100$ a
| month.
| highwaylights wrote:
| I'm not seeing the relevance of YouTube and the One services
| to this _at all_.
|
| I get that Big Tech loves to try to pull you into their orbit
| whenever you use one of their services, but this risks
| alienating customers who won't use those unrelated services
| and may begrudge Google making them pay for them.
| j_maffe wrote:
| Idk if anyone will see these offerings more than just an
| added bonus, especially when you compare to OAI which asks
| for more for only the AI models.
| OJFord wrote:
| It's trying to normalise it, make it just another part of
| your Google experience, alongside (and integrated with)
| your other Google tools. (Though that's weakened a bit by
| calling it ' _AI_ Pro /Ultra' imo.)
| bezier-curve wrote:
| For $250/mo I would hope it includes API access to Gemini 2.5
| pro, but it's nice to want things.
| highwaylights wrote:
| I can't see a way that anyone would be able to give uncapped
| access to these models for a fixed price (unless you mean it's
| scoped to your own use and not for company use? Even then,
| that's still a risk to the provider.)
| bezier-curve wrote:
| I use Msty a lot for personal use. I like its ability to fork
| conversations. Seems like a simple feature but even ChatGPT's
| UI, which everyone has tried to copy, is fairly limited by
| comparison.
| pc86 wrote:
| As a consumer it seems to me the low hanging fruit for these
| super-premium offerings is some substantial amount of API
| credits included every month. Back when API credits were a
| relatively new thing I used LLMs infrequently enough I just
| paid $5-10/mo for API credits and used a desktop UI to talk to
| ChatGPT.
|
| Now they want $200, $250/mo which is borderline offensive, and
| you have to pay for _any_ API use on top of that?
| Aurornis wrote:
| Putting API use into the monthly plans doesn't make a lot of
| business sense. The only people who would sign up specifically
| to use API requests on a monthly plan would be looking to have
| a lower overall bill, then they'd pay-per-request after that.
| It would be a net loss.
| mrbluecoat wrote:
| Amusing they picked a deadly jellyfish attacking earth as their
| hero image
| sharpshadow wrote:
| Technically more than one could use this subscription, if used
| through the same device. Also to make it available to users in
| not yet supported countries.
| aylmao wrote:
| I can't decide how I feel about Google's design for this looking
| so Apple-y.
|
| Didn't they just release Material Design Expressive a few days
| ago [1]? Instead of bold shapes, bold fonts and solid colors it
| gradients, simple lines, frosted glass, and a single, clean,
| sans-serif font here. The bento-box slides look quite Apple-y too
| [2]. Switch the Google Sans for SF Pro, pull back on the border
| radius a bit, and you've essentially got the Apple look. It does
| look great though.
|
| [1]: https://news.ycombinator.com/item?id=43975352
|
| [2]: https://blog.google/products/gemini/gemini-app-updates-
| io-20...
| GuinansEyebrows wrote:
| it makes sense if you believe that Google has zero business
| interest in UI/UX.
|
| they've learned that they can shovel out pretty much anything
| and as long as they don't directly charge the end-user and
| they're able to put ads on it (or otherwise monetize it against
| the interest of the end user), they just don't care.
|
| they've been criticized for years and years over their lack of
| standardization and relatively poorly-informed design choices
| especially when compared with Apple's HIG.
| solomatov wrote:
| Is there a premium option to control which of your data will be
| used for training? Or is it implemented the same way as Gemini
| Pro?
| rudedogg wrote:
| If anyone at Google cares, I'd pay for Gemini Pro (not this new
| $200+ ridiculousness) if they didn't train on my chats in that
| case. I actually would _like_ to subscribe..
| j_maffe wrote:
| There's already an option for that. The downside is you can't
| access your chat history.
| buildfocus wrote:
| And lots of other features don't work, particularly
| external integrations. Gemini on Android refuses to do
| basic things like set a timer unless chat history is
| enabled. It is the one key feature I really want to pay
| extra to get, and that preference goes x2 when the AI
| provider is Google.
| submeta wrote:
| Google Ultra: USD 250. Claude Pro: 218 EUR ChatGPT Pro: 220 EUR
|
| Not included: Perplexity, Openrouter, Cursor, etc
|
| Wow. You gotta have lots of disposable income.
| thenaturalist wrote:
| There are enough people who do. ;)
|
| And from a business perspective, this is enabling people from
| solo freelancers to mid managers and others for a fraction of
| the time and cost required to outsource to humans.
|
| Not that I am personally in favor of this, but I can very much
| see the economics in these offerings.
| loudmax wrote:
| The target market for these offerings are corporations, or
| self-employed developers. If these tools really do make your
| developers more productive, $250 a month is easily justifiable.
| That's only $3000 per year, a fraction of the cost of a full
| time developer.
|
| Obviously, the benefit is contingent on whether or not the
| models actually make your developers more productive.
| catigula wrote:
| Do any of these companies actually sell their products as
| developer replacements?
| mwigdahl wrote:
| No, but they do sell them as developer augmentations.
| 6510 wrote:
| I hear it doesn't have to be productive right now, if you
| have deep pockets it is worth being familiar with the tools
| even if it is just in case.
| mattfrommars wrote:
| $250 a month is still a lot of money in India to spend on a
| digital product.
| kkarakk wrote:
| AI has completely eliminated salary adjusted pricing in
| software. no discounts for 3rd world in anything.
| ivm wrote:
| Yup, I was paying $225/mo for three Unity3D subscriptions
| (basic+iOS+Android) ten years ago, while earning less than
| $4k/mo - just considered it part of my self-employed
| expenses.
| eastbound wrote:
| The goal is to capture all your disposable AI income, so they
| can starve the competitors. The goal is, as long as you
| subscribe to several, increase the price.
|
| And you haven't strung the price that stings yet.
| ZeroTalent wrote:
| that $250/month can make you $20k/month if you do some
| automation and subjectively unethical things
| lazharichir wrote:
| Like taking on a gazillion of contract work?
| mirkodrummer wrote:
| Apart from clients that usually are not stupid, you still
| need to understand requirements to guide the ai in a
| possible right direction. I don't usually understand boss
| tickets at first look, very often we need to discuss them,
| i doubt an ai could despite the hype
| add-sub-mul-div wrote:
| And this is still early days, the pre-enshittification era.
| paxys wrote:
| This isn't a luxury purchase. If you aren't able to increase
| your income by $250/mo using these tools or otherwise get
| $250/mo worth of value out of them then you shouldn't sign up.
| i_love_retros wrote:
| This is all a bit silly
| qweiopqweiop wrote:
| Am I the only one getting the AI fatigue?
| hammock wrote:
| Don't worry, everyone is. That is why they are switching to
| quantum soon
| charles_f wrote:
| Quantum AI on the blockchain, now in Rust.
| spookie wrote:
| With data analytics visualized in AR
| tsujamin wrote:
| Obviously not speaking for others experience, but it all makes
| me feel pretty fatigued, and as if this growing expectation of
| "AI-enhanced productivity" is coming at the expense of a craft
| and process (writing software) that I enjoy.
| zhivota wrote:
| Ok so I have Google One AI or whatever the previous version of
| this is called, and what's wild to me is that in Google Sheets,
| if I ask it to do anything, literally anything, it says it can't
| do it. The only thing it can do is read a sheet's data and talk
| about it. It can't help you form formulas, add data to the sheet,
| organize things, anything, as far as I've seen.
|
| How does Google have the best models according to benchmarks but
| it can't do anything useful with them? Sheets with AI assist on
| things like pivot tables would be absolutely incredible.
| noosphr wrote:
| >How does Google have the best models according to benchmarks
| but it can't do anything useful with them?
|
| KPI driven development with no interest in killing their cash
| cow.
|
| These are the people who sat on transformers for 5 years
| because they were too afraid it would eat their core business,
| e.g. Bert.
|
| One need look at what Bell Labs did to magnetic storage to
| realize that a monopoly isn't good for research. In short: we
| could have had mass magnetic storage in the 1920s/30s instead
| of 50s/60s.
|
| A pop sci article about it: https://gizmodo.com/how-ma-bell-
| shelved-the-future-for-60-ye...
| catigula wrote:
| I mean there's an article in Fortune magazine about the
| people pushing transformer "research" building doomsday
| bunkers.
|
| Making Google look like the mature person in the room is a
| tall order but it seems to have been filled.
| wrs wrote:
| The AI assistant in Sheets doesn't understand how even the
| basic features work. When it's not saying it lacks basic
| knowledge, it hallucinates controls that don't exist. Why even
| bother having it there?
| zb3 wrote:
| Good good, please buy it so I can continue using Gemini for free.
| retep_kram wrote:
| Given that the AI scene is not stable at all in the current
| moment (every day a new release that make last month's obsolete),
| any offer that tries to lock you with a model or model provider
| is a bad idea.
|
| Pay-per-use for the moment, until market consolidation and/or
| commoditization.
| Aurornis wrote:
| > any offer that tries to lock you with a model or model
| provider is a bad idea.
|
| It's a monthly plan that you can cancel at any time. Not really
| locking in.
| sandspar wrote:
| 30TB of Google storage is a soft lock-in. If you fill it up
| you're kinda stuck.
| charles_f wrote:
| This is the kind of pricing that I expect most AI companies are
| gonna try to push for, and it might get even more expensive with
| time. When you see the delta between what's currently being burnt
| by OpenAI and what they bring home, the sweet point is going to
| be hard to find.
|
| Whether you find that you get $250 worth out of that subscription
| is going to be the big question
| Ancapistani wrote:
| I agree, and the problem is that "value" != "utilization".
|
| It costs the provider the same whether the user is asking for
| advice on changing a recipe or building a comprehensive project
| plan for a major software product - but the latter provides
| much more value than the former.
|
| How can you extract an optimal price from the high-value use
| cases without making it prohibitively expensive for the low-
| value ones?
|
| Worse, the "low-value" use cases likely influence public
| perception a great deal. If you drive the general public off
| your platform in an attempt to extract value from the
| professionals, your platform may never grow to the point that
| the professionals hear about it in the first place.
| garrickvanburen wrote:
| this is the problem Google search originally had.
|
| They successfully solved it with an advertising....and they
| also had the ability to cache results.
| mysterydip wrote:
| Do LLMs cache results now? I assume a lot of the same
| questions get asked, although the answer could depend on
| previous conversational context.
| jsheard wrote:
| I wonder who will be the first to bite the bullet and try
| charging different rates for LLM inference depending on
| whether it's for commercial purposes. Enforcement would be a
| nightmare but they'd probably try to throw AI at that as
| well, successfully or not.
| chis wrote:
| I think there are always creative ways to differentiate the
| two tiers for those who care.
|
| "Free tier users relinquish all rights to their
| (anonymized) queries, which may be used for training
| purposes. Enterprise tier, for $200/mo, guarantees queries
| can only be seen by the user"
| emzo wrote:
| This would be great for open source projects
| jfrbfbreudh wrote:
| This is what Google currently does for access to their
| top models.
|
| AI Studio (web UI, free, will train on your data) vs API
| (won't train on your data).
| typewithrhythm wrote:
| Value capture pricing is a fantasy often spouted by salesmen,
| the current era AI systems have limited differentiation, so
| the final cost will trend towards the cost to run the system.
|
| So far I have not been convinced that any particular platform
| is more than 3 months ahead of the competition.
| morkalork wrote:
| Costs more than seats for Office 365, Salesforce and many
| productivity tools. I don't see management gleefully running to
| give access to whole departments. But then again, if you could
| drop headcount by just 1 on a team by giving it to the rest,
| you probably come out ahead.
| EasyMark wrote:
| I feel prices will come down a lot for "viable" AI, not
| everyone needs the latest and greatest at rock-bottom prices.
| Assuming AGI is just a pipe-dream with LLMs as I suspect.
| Wowfunhappy wrote:
| > When you see the delta between what's currently being burnt
| by OpenAI and what they bring home, the sweet point is going to
| be hard to find.
|
| Moore's law should help as well, shouldn't it? GPUs will keep
| getting cheaper.
|
| Unless the models also get more GPU hungry, but 2025-level
| performance, at least, shouldn't get more expensive.
| dvt wrote:
| > Moore's law should help as well, shouldn't it? GPUs will
| keep getting cheaper.
|
| Maybe I'm misremembering, but I thought Moore's law doesn't
| apply to GPUs?
| godelski wrote:
| Not necessarily. The prevailing paradigm is that performance
| scales with size (of data and compute power).
|
| Of course, this is observably false as we have a long list of
| smaller models that require fewer resources to train and/or
| deploy with equal or better performance than larger ones.
| That's without using distillation, reduced
| precision/quantization, pruning, or similar techniques[0].
|
| The real thing we need is more investment into reducing
| computational resources to train and deploy models and to do
| model optimization (best example being Llama CPP). I can tell
| you from personal experience that there is much lower
| interest in this type of research and I've seen plenty of
| works rejected because "why train a small model when you can
| just tune a large one?" or "does this scale?"[1] I'd also
| argue that this is important because there's not infinite
| data nor compute.
|
| [0] https://arxiv.org/abs/2407.05694
|
| [1] Those works will out perform the larger models. The
| question is good, but this creates a barrier to funding.
| Costs a lot to test at scale, you can't get funding if you
| don't have good evidence, and it often won't be considered
| evidence if it isn't published. There's always more
| questions, every work is limited, but smaller compute works
| have higher bars than big compute works.
| jorvi wrote:
| Small models will get really hot once they start hitting
| good accuracy & speed on 16GB phones and laptops.
| fellowniusmonk wrote:
| As someone who grew up very poor and first got access to email
| via Juno's free ad based email client.
|
| I've seen so many people over the years just absolutely shit on
| ad based models.
|
| But ad based models are probably the least regressive approach to
| commercial offerings that we've seen work in the wild.
|
| I love ads. If you are smart you don't have to see them. If you
| are poor and smart you get free services without ads so you don't
| fall behind.
|
| I notice that there are no free open source providers of LLM
| services at this point, it's almost as if services that have high
| compute costs have to be paid for SOME HOW.
|
| Hopefully we get a Juno for LLM soon so that whole cycle can
| start again.
| leoh wrote:
| Ads have really harmed our society imo, despite having some
| advantages as you mention
| danesparza wrote:
| I don't mean to be snarky, but is this announcement timed just so
| they take press away from the Microsoft Copilot announcement?
| jonas21 wrote:
| Today was the Google I/O keynote. The date was set months in
| advance.
| Analemma_ wrote:
| It's exactly the opposite: the date for I/O was fixed months
| ago, Microsoft made their announcement to try and take press
| away from Google.
| Ancapistani wrote:
| I've toyed with Gemini 2.5 briefly and was impressed... but I
| just can't bring myself to see Google as an option as an
| inference provider. I don't trust them.
|
| Actually, that's not true. I do trust them - I trust them to
| collect as much data as possible and to exploit those data to the
| greatest extent they can.
|
| I'm deep enough into AI that what I really want is a personal RAG
| service that exposes itself to an arbitrary model at runtime. I'd
| prefer to run inference locally, but that's not yet practical for
| what I want it to do, so I use privacy-oriented services like
| Venice.ai where I can. When there's no other reasonable
| alternative I'll use Anthropic or OpenAI.
|
| I don't trust any of the big providers, but I'm realizing that I
| have baseline hostility toward Google in 2025.
| nowittyusername wrote:
| Understanding that no outside provider is going to care about
| your privacy and will always choose to try and sell you their
| crappy advertisements and push their agenda on you is the first
| step in building a solution. In my opinion that solution will
| come in the form of a personalized local ai agent which is the
| gatekeeper of all information the user receives and sends to
| the outside world. A fully context aware agent that has the
| users interests in mind and so only provides user agreed
| context to other ai systems and also filters all information
| coming to the user from spam, agenda manipulation, etc...
| Basically a very advanced spam blocker of the future that is
| 100% local and fully user controlled and calibrated. I think we
| should all be either working on something like this if we want
| to keep our sanity in this brave new world.
| Ancapistani wrote:
| Exactly.
|
| To be clear, I don't trust Venice either . It just seems less
| likely to me that they would both lie about their collection
| practices and be able to deeply exploit the data.
|
| I definitely want locally-managed data at the very least.
| CSMastermind wrote:
| I pay for OpenAI Pro but this is a clear no for me. I just don't
| get enough value out of Gemini to justify a bump from $20 / month
| to $250.
|
| If they really want to win they should undercut OpenAI and
| convince people to switch. For $100 / month I'd downgrade my
| OpenAI Pro subscription and switch to Gemini Ultra.
| radicality wrote:
| It does look like it comes with a few other perks that would
| normally cost a bunch too, specifically, 30TB of Google drive
| storage
| airstrike wrote:
| Yeah, no, thanks for the cross-sell but I'm not interested.
| J_Shelby_J wrote:
| If they really want to win, they should make a competitor for
| O1-pro. It's worth $200 to reduce LLM babysitting needs by %10.
| mvdtnz wrote:
| Perhaps they're not interested in beating OpenAI in the
| business of selling $1 for $0.50.
| CSMastermind wrote:
| Sure but failure to capture marketshare now could easily
| snowball into larger failure of the business later.
| hackrmn wrote:
| Ok, this reminds me of that Black Mirror episode,
|
| *spoilers ahead*
|
| where the lady had a fatal tumor cut out for emergency procedure,
| only for it to be replaced by a synthetic neural network used by
| a cloud service with a multi-tier subscription model where even
| the basic features are "conveniently" shoved into a paying tier,
| up until the point she's on life support after being unable to
| afford even the basic subscription.
|
| Life imitates art.
| j_maffe wrote:
| wdym life imitates art? This is exactly what the episode was
| about lol
| johnisgood wrote:
| Art imitates life!
| OtherShrezzing wrote:
| The global average salary is somewhere in the region of $1500.
|
| There's lots of people and companies out there with $250 to spend
| on these subscriptions per seat, but on a global scale (where
| Google operates), these are pretty niche markets being targeted.
| That doesn't align well with the multiple trillions of dollars in
| increased market cap we've seen over the last few years at
| Google, Nvda, MS etc.
| paxys wrote:
| New technology always starts off available to the elite and
| then slowly makes its way down to everyone. AI is no different.
| dimitrios1 wrote:
| This is one of those assumed truisms that turns out to be
| false upon close scrutiny, and there's a bit of survivorship
| bias in the sense that we tend to look at the technologies
| that had mass appeal and market forces to make them cheaper
| and available to all. But theres tons of new tech thats
| effectively unobtainable to the vast majority of populations,
| heck even nation states. With the current prohibitive costs
| (in terms of processing power, energy costs, data center
| costs) to train these next generation models, and the walled
| gardens that have been erected, there's no reason to believe
| the good stuff is going to get cheaper anytime soon, in my
| opinion.
| paxys wrote:
| > turns out to be false upon close scrutiny
|
| Care to share that scrutiny?
|
| Computers, internet, cell phones, smartphones, cameras,
| long distance communication, GPS, televisions, radios,
| refrigerators, cars, air travel, light bulbs, guns, books.
| Go back as far as you want and this still holds true. You
| think the the majority of the planet could afford any of
| these on day 1?
| kkarakk wrote:
| the point is not that AI services will be affordable
| "eventually" it's that the advantage is so crazy that
| people who don't have access to them will NEVER be able
| to catch up. First AI wrappers disrupt industries
| ->developing nations can't compete coz the services are
| priced prohibitively -> AI wrappers take over even more
| -> automation disrupts the need for anyone -> developing
| nations never develop further. this seems more and more
| likely not less. cutting edge GPUs for eg - already are
| going into the stratosphere pricing wise and are
| additionally being sanctioned off.
| tekla wrote:
| How is this different from literally all of human history
| hombre_fatal wrote:
| It seems you're suggesting that once you start this
| process of building tech on top of tech, then you get far
| ahead of everyone because they all have to independently
| figure out all the intermediate steps. But in reality,
| don't they get to skip to the end?
|
| e.g. Nations who developed internet infrastructure later
| got to skip copper cables and go straight to optical tech
| while US is still left with old first-mover
| infrastructure.
|
| AI doesn't seem unique.
| sxg wrote:
| I disagree. There are massive fixed costs to developing
| LLMs that are best amortized over a massive number of
| users. So there's an incentive to make the cost as cheap as
| possible and LLMs more accessible to recoup those fixed
| costs.
|
| Yes, there are also high variable costs involved, so
| there's also a floor to how cheap they can get today.
| However, hardware will continue to get cheaper and more
| powerful while users can still massively benefit from the
| current generation of LLMs. So it is possible for these
| products to become overall cheaper and more accessible
| using low-end future hardware with current generation LLMs.
| I think Llama 4 running on a future RTX 7060 in 2029 could
| be served at a pretty low cost while still providing a ton
| of value for most users.
| TulliusCicero wrote:
| Yeah, GP is overextending by saying it's always true.
|
| The more basic assertion would be: something being
| expensive doesn't mean it can't be cheap later, as many
| popular and affordable consumer products today started out
| very expensive.
| pier25 wrote:
| Do you have a source for the $1500 number? Seems pretty high.
| bradleybuda wrote:
| It's 6x that. The median is 2x:
| https://chatgpt.com/share/682ceb2a-b56c-800b-b49d-1a24c48709...
| julianpye wrote:
| Why do people keep on saying that corporations will pay these
| price-tags? Most corporations really keep a very tight lid on
| their software license costs. A $250 license will be only
| provided for individuals with very high justification barriers
| and the resulting envy effects will be a horror for HR. I think
| it will be rather individuals who will be paying out of their
| pocket and boosting their internal results. And outside of those
| areas in California where apples cost $5 in the supermarket I
| don't see many individuals capable of paying these rates.
| troupo wrote:
| Corps will likely negotiate bulk pricing and discounts, with
| extra layers of guarantees like "don't use and share our data"
| on top
| bryanlarsen wrote:
| "AI will make us X% more productive. 100%-X% of you are fired,
| the rest get a $250/month license".
| kulahan wrote:
| I don't see any benefit to removing humans in order to
| achieve the exact same level of efficiency... wouldn't that
| just straight-up guarantee a worse product unless your
| employees were absolutely all horrendous to begin with?
| verdverm wrote:
| We just signed up to spend $60+/month for every dev to have
| access to Copilot because the ROI is there. If $250/month save
| several hours per month for a person, it makes financial sense
| delusional wrote:
| We signed up for that too. 2 quaters later the will to pay is
| significantly lower.
| tacker2000 wrote:
| How are you measuring this? How do you know it is paying off?
| afroboy wrote:
| And why AI hype train didn't work on gaming industry? why
| it didn't save hundreds of hours from game devs times to
| get latest GTA anytime sooner?
|
| I'm not sure it's correct that we need to measure the
| benefits of AI depending on the lines of codes that we
| wrote but on how much we ship more quality features faster.
| julianpye wrote:
| Okay, but you're in a S/W team in a corp, where everyone's
| main task is to code. A coding agent has clear benefits here.
|
| This is not the usecase of AI Ultra.
| ir77 wrote:
| people here keep saying that this is targeted at big
| companies/corporations. the big company that i work for
| explicitly block uploads of data to these services and we're
| forbidden to put anything company related in there for many
| reasons, even if you use your own account, we don't have 'company
| accounts'.
|
| so no, i can't see companies getting all excited about buying
| $250mo/user licenses for their employees for google or chatgpt to
| suck in their proprietary data.
| verdverm wrote:
| These subscriptions explicitly do not suck in your proprietary
| data, it's all laid out in their ToS.
| quantumHazer wrote:
| Yeah, and who will make them accountable? How can you verify
| that they're noteworthy stealing your data anyways? This
| companies don't give a shit about copyright or privacy.
| ndriscoll wrote:
| I don't think I've ever worked somewhere where you _wouldn
| 't_ get fired for sending company data to a party that
| doesn't have an NDA signed with the company, regardless of
| whatever terms they told you personally.
| sigmaisaletter wrote:
| The same companies who stole... sorry.. fair-used all the
| worlds artworks and all text on the internet to train their
| models are now promising you they won't steal...sorry...
| fair-use your uploaded data?
|
| In unrelated matters, I have a bridge to sell you, if you are
| interested.
| spaceman_2020 wrote:
| Honestly, at this point, nation states will have to figure out an
| AI strategy. Poorer countries where the locals can't afford
| cutting edge AI tools will find themselves outproduced by
| wealthier workers.
| dbspin wrote:
| Not just can't afford, but can't access. Most of these new AI
| tools aren't available outside the US.
| Papazsazsa wrote:
| I just tried it.
|
| 8 out of 10 attempts failed to produce audio, and of those only 1
| didn't suck.
|
| I suppose that's normal(?) but I won't be paying this much
| monthly if the results aren't better, or at least I'd expect some
| sort of refund mechanism.
| johnisgood wrote:
| They are running out of ideas for names. What next, Google AI
| Ultra Max Pro?
| loloquwowndueo wrote:
| You joke but don't forget there's an actual product called
| "product name" pro max (an iPhone lol)
| Fergusonb wrote:
| That's a car payment...
| backendEngineer wrote:
| All to put even more AI generated bs down our throats to sell us
| trash that we don't need and can't afford. God the roaring
| twenties can't catch a break.
| MOARDONGZPLZ wrote:
| I use it to make meeting agendas.
| icelancer wrote:
| I paid for it and Google Flow was upgraded to Ultra but Gemini
| still shows Pro, and asks for me to upgrade. When I go to
| "upgrade," it says I am already at Google Ultra.
|
| Average Google product launch.
___________________________________________________________________
(page generated 2025-05-20 23:00 UTC)