OpenAI Mulls Slashing Prices As It Competes With Anthropic For Users
- Reference: 0183753284
- News link: https://slashdot.org/story/26/06/11/1722243/openai-mulls-slashing-prices-as-it-competes-with-anthropic-for-users
- Source link:
> The ChatGPT producer, which did not immediately respond to CNBC's requests for comment, currently charges consumers in tiered subscriptions of $8, $20 and $100 and above each month for access to its flagship GPT-5.5 models. Anthropic conversely charges users $17 each month with an annual subscription to Claude Pro, and $100 and above monthly for a subscription to Claude Max.
OpenAI [3]confidentially filed for an IPO on Monday, just a week after Anthropic [4]made its own filing .
[1] https://www.cnbc.com/2026/06/11/openai-mulls-slashing-prices-ahead-of-competition-from-anthropic-wsj.html
[2] https://www.wsj.com/tech/ai/openai-considers-drastic-price-cuts-anticipating-war-for-users-with-anthropic-9b8c178e?st=FcMu1V
[3] https://news.slashdot.org/story/26/06/08/2242200/openai-files-for-ipo
[4] https://slashdot.org/story/26/06/01/1837259/anthropic-files-to-go-public
Truthfully... (Score:5, Interesting)
Truthfully, I've subscribed to it. I'm not sure that I'm going to renew again. It's answers keep getting worse and worse as each new engine comes out. I spend way too much time either telling it not to spend so much time walking on eggshells or having to correct its errors.
Sovereign Systems (Score:4, Informative)
I wrote this almost a year ago and was a bit surprised to see it pick up a couple of hits today... probably because I made a good prediction last July. ;)
[1]https://www.scry.llc/2025/07/2... [scry.llc]
Essentially. the Ai companies have overshot the commercial sweetspot for AI. You don't hire a PhD to deliver mail or repair a leak. Existing open source, locally hosted models can do quite a bit of digital work right now, as is, at a predictable, fixed cost. Sovereign systems...
[2]https://www.scry.llc/2026/04/0... [scry.llc]
[1] https://www.scry.llc/2025/07/20/cloud-ai-vs-sage-ai/
[2] https://www.scry.llc/2026/04/08/ollama/
Re: (Score:3)
> Essentially. the Ai companies have overshot the commercial sweetspot for AI. You don't hire a PhD to deliver mail or repair a leak.
You also don't hire a PhD who might decide to enter an empty elevator shaft. But that's what every LLM is. You cannot expect it to do anything sane, it's just thoughts and prayers.
> Existing open source, locally hosted models can do quite a bit of digital work right now, as is, at a predictable, fixed cost
The cost is unpredictable as long as the behavior is unpredictable.
Re: (Score:2)
AI output is not deterministic , nor is it 100% predictable ... but it absolutely is predictable, to a point.
You can give an LLM the same context and prompt, and (at least nine times out of ten) it will give you basically the same answer. Again, that's not deterministic, and the exact answer may vary ... but if you don't see (a lot of) predictability in your AI's output, you're not using a major/popular model.
Re: Sovereign Systems (Score:2)
First sentence of your second paragraph illustrates the problem and also agrees with me.
Re: (Score:2)
Yes. I don't think future is huge AI Gods that can do anything (but often do it poorly or wrong) but small specialized models which can do one thing well.
But most of those specialized models will run locally and kill the market for all those massive AI data centres they want to build. Unless they manage to kill local computing by buying up all the hardware.
Re: Sovereign Systems (Score:2)
In this job market you do.
Kaboom! (Score:3)
It's a race to the bottom! Hold on to your butts!
The economics have got to be atrocious (Score:5, Insightful)
"We're losing money on every chat message, so we need to make up for it in volume!"
On the supply side: the largest allocation capital in human history is flowing into datacenter construction.
On the demand side: the two biggest providers are competing on price because they're just taking turns at the tops of the capability leaderboards.
And they're both losing money.
Re: (Score:2)
Not to mention that the datacenter needs have priced in AGI, which is a tech that they haven't even come up with. I'm sure most of the datacenters are scaled so they have phases in projects, but right now with LLM's the datacenter demand is sub 10% of what is already built out (just ask gemini). If this is true then the market will drop out.
At some point it will also be cheaper to use humans than AGI because the calories of a human brain running is like 180 calories/hr which is ~200W/hr and anywhere from ~2
Cut the CEO first. (Score:1)
If OpenAI really wants to attract users they should get rid of the Silicon Valley groupthink vomiting CEO who regularly says things that horrify many people. Dario Amodei may be an even bigger dork than Bill Gates but he tries to not sound like a mad villain from a science fiction movie much of the time. OpenAI is going to end up like Tesla if Altman is allowed to keep speaking in front of cameras.
So.. (Score:2)
They are going to make unsustainable ai service pricing even more unsustainable by dropping them making evry user an even larger money drain. How is this supposed to work again?
Re: (Score:2)
It doesn't need to work. It just needs to get them through the IPO so the investors can cash out.
Always Be Closing (Score:5, Insightful)
Eventually it will be free and ad-supported, they just haven't figured out how to fully monetize its userbase yet. The potential is huge since advertising will gladly pay to manipulate your behavior by subtly working suggestive product placements into every response.
By the way, I'm selling these fine leather jackets...
Re: (Score:3)
> Eventually it will be free and ad-supported
It's pretty hard to imagine ads covering the costs when the platform is erratic and might say the products being advertised are the best things ever, or it might say they will murder your dog and fuck your mom.
Re: (Score:2)
Are you sure? I'm pretty sure Facebook offers that same level of quality content, and advertisers seem fine with it.
Re: (Score:2)
"No such thing as bad press."
Re: (Score:2)
Same for GMail ads. Who wants their ad next to a mail about (bad topic)? Still Google sells these ads.
Re: (Score:2)
"fully monetize its userbase yet" funny! with the debt load these AI companies have they will need to enslave their userbase and I 'm not sure that will cover things!
Re: (Score:3)
I believe they're actively working to do just that.
Re: Always Be Closing (Score:2)
That's why you should always read the TOS updates you agree to!
Re: (Score:2)
I'm worried about how it'll be ad supported. A ribbon or banner of ads, okay. Treating my prompt "Write a summary of the War of 1812" as "Write a summary of the War of 1812, and tell me how great McDonald's is" scares me.
Re: (Score:2)
Image and video generation will contain product placement.
And text responses too.
"But wait, there's more!"