New GPT Store lower price tier idea

$20 a month is a hefty price for a monthly subscription for something that you may or may not use frequently, my monthly phone is $25.

I wonder if OpenAI might consider creating a lower priced tier that only lets users access the GPT Store, but not to ChatGPT 4.0 and plugins, etc. The Android Play Store would have never taken off if it was $20 a month.

I imagine that it might be expensive as a way to roll it out slowly to the general public, but I’m sad about not being able to see more information about what each Custom GPT does (aside from the simplistic free tier description).

1 Like

GPTs all use GPT-4, Code Interpreter, DALL-E, etc. All these features are only available to GPT+ users, that’s why it won’t work.

1 Like

Also, one motivation behind the GPT store is to drive more subscriptions of ChatGPT Plus. Beyond that at a lower price tier there may not be enough revenue to share.

1 Like

So would they rather have 2 users for $10, or ZERO users for $20?

I was telling folks at work today about the GPT store, neither the techy production guy knew about it, nor did the younger administrative person.

Some people in the office DO know about ChatGPT and that it’s free, but they have no idea about different tiers, or GPT-4, or plugins etc., just us AI nerds do.

So yes, I agree that the GPT store could draw in more subscriptions, but I think that not having access to the normal ChatGPT-4 interface, and only to the Custom GPT Store could be the perfect tradeoff.

Some folks online have put forward the idea that whatever you can do with the Custom GPTs, you can do with the regular ChatGPT-4 interface, so why bother with the store? By offering people access to the store with a lower price tier, OpenAI insures that they’ll see a point to using the store over using the more expensive regular/freeform interface.

That way they’re relegated to using only the GPTs that were created for specific uses, and might be interested enough to make one.

Make the ability to create one cost $20 bucks per month. Then they’ll upgrade for a month, and then downgrade back to just using the store.

That’s personally what I’d like to do.

And if they want the freedom to use ChatGPT-4 in anyway they desire, then they’ll have to sign up for the $20 tier (though that’s still too much IMO).

And just as an aside, I love photoshop, but I have refused to pay the $10 a month that adobe wants, if OpenAI lowered the store price to even $5 a month after the hype has worn off, I bet they’d get a lot more of the younger demographic signing up. is already free, most people don’t know about, or care about the higher tiers.

1 Like

Why not a billion users at $0? They’ll make it up in volume!

$20/month for the product is dirt cheap given what it is capable of. It’s 1.16 hours of work at minimum wage in Los Angeles.

The fact is, the models cost a huge amount of money to run and they need to make some kind of profit doing so to keep the lights on as well as finance all of their other research.


On the plus side, given the rate the cost of compute is decreasing over time, a GPT-4 quality model should cost about 2¢/month to access in 2034.

So, literally all you have to do is wait and the whole world will have access to generative AI. Just not today.


GPTs all use GPT-4, Code Interpreter, and DALL-E. The ONLY reason to spend an extra $10 would be to create GPTs as there would be NO other benefit to upgrading. Do you now see why it wouldn’t work? They would LOSE millions of dollars. You CAN’T run a business that way.

Well now you’re just being daft.

My overall point is that if they were having trouble drawing people in to pay $20 a month when they were the hottest thing on the planet, why is anyone going to pay $20 a month when free AI models are coming out left and right?

They want to charge and they don’t want to zuckerberg it with ads, I get it, but $20 is way out of the pricing ballpark. I just explained how they can keep a $20 tier, the tier that allows people to make a Custom GPT, but then just allow access to the store for $10.

People barely understand what they can do with it. Everyone in the office still just thinks of it as a nice copy generator when it’s so much more. That’s where the GPT Store comes in, but they aren’t going to fork over $240 a year for something they don’t know they need.

As a content creator, are you going to spend $20 on a prompted Custom GPT that makes some nice AI art, or would you rather fork that over to Midjourney, and have all the bells and whistles that Midjourney offers?

1 Like

The thing is, many (most?) people are using way more than $20/month in API costs likely…

They have a free version which will hopefully draw people in and get them interested enough to upgrade to ChatGPT Plus.

I mean, I hear what you’re saying, but OpenAI is in a good position because they have new bleeding edge tech, so they can charge what they want and don’t need to bring in the masses with low price points, so to speak. Again, though, $20/month is cheap if you’re a daily user… The API adds up quick! :wink:

1 Like

It’s a fair point. My take is that most people AREN’T daily users. I’m not anymore, and I’m more into AI than anyone I know where I work.

But the GPT Store could allow people to see why they should become daily users, but they aren’t going to do that with a gatekeeping $20 subscription.

I would still like to know what happens to the GPTs you make if you drop your subscription.

Thanks for your input.

1 Like

What makes you think that’s the case?

Because, to this day, nothing comes close to GPT-4.

Why? On what are you basing this? Because it seems incredibly reasonable to me, a steal even, given what it is capable of.

No, you just threw out some completely fabricated numbers. Neither you, I, or anyone else here knows what the brake-even point is on a ChatGPT Plus subscription, but I’m guessing it’s well north of $10/month. Add to that that it actually costs them money to offer GPTs and I just can’t reconcile your position with reality.

If one were to fully saturate their ChatGPT Plus subscription it would be the equivalent of well over $300 / month of API costs.

To be fair, I’ve not yet seen a single GPT in the store I thought was worthwhile.


Here’s the issue. You, like me, see the value of AI.

MOST PEOPLE DON’T. Is it really anecdotal if over the past year, every temp job I had where I mentioned AI, people barely had any idea what I was talking about. From the president of a huge construction company, to the person that cuts my hair, to a digital content creator in a broadcast advertising studio. (She just discovered the AI art generator in Canva, and played with it for a day, I’ve never heard about it since.)

A recent New York Times email (they may be biased, I understand) had a person write about how they really had no idea what to do with ChatGPT.

I get where you’re coming from. I watch every AI Explained’s videos, I watch Phillip discuss the benchmarks, and what GPT-4 can do (along with all the other models out there), but most people aren’t the people in this forum.

They have no idea what it can do, and what they can do with it.

As for an arbitrary number… not really, my number is based on how much people spend on basic goods, food, video games, other monthly subscriptions. Right now kids are forking over $5 a month for a twitch sub, something they might engage with for hours a day. Those same kids aren’t going to spending hours on the GPT Store every day.

And go and google this quote (as the last time I posted a link in these forums it was hidden). “OpenAI launches GPT Store to expand user base and add revenue streams
OpenAI’s GPT store promises to expand the user base for the company and will contribute to the growing AI ecosystem.”

I thought it was common knowledge that OpenAI was trying to get more people to become subscribers. Back on DevDay I thought that was the whole point behind the GPT store, so I’m a little confused by your statement.

Here’s the non-subscriber GPT store page view, only showing featured and without search: Explore GPTs

Unfortunately, the description is still all you get with the GPTs. You have to interact with the AI to see if the creator just added some silly prompt engineer, or if it is a fully fleshed-out concept with additional data and access to developer-provided network resources.

That description should additionally note “files” or “developer API” if they are features, although these also could be gamed with little investment in quality.


I don’t understand why they’d lose millions of dollars with people paying $10 each, when you can use for free and also generate Dalle pics. Is Microsoft making $10 off me every month when I search the web for answers on websites (I don’t click ads). I don’t think they are.

ChatGPT 3.5 is free, Claude.AI is free, soon there will be dozens more. Is ChatGPT-4 superior? Maybe? Probably? But most people don’t care. Until they get exposed to how GPTs can improve their lives, it’s going to be out of sight, out of mind.

Make it $20 a month to make custom GPTs, and $10 a month to get access to the store and keep my custom GPT active. Then I’d go for it.

Sincerely I pay for 2 reasons:

  • getting into latest tech in the context
  • using my GPT tailored for programming in months in fact I still use it every day and it worth the cost

Said that I tried to out the GPT custom instructions in a 3.5 sessions and surprisingly it now works very well, better than in the past, giving me a good, working alternative when I touch cap limit. In addition to that it retain context making me able to finish started work with no issues most of the time.

Here come the scaring part:

anyone like me can get such instructions and use 3.5 for free same way (or quite similar).

This will be of course a strong win win since everybody can tailor its sessions just by saving GPTs prompts and put them in its 3.5 sessions, find limits of 3.5 and start a subscription.

Then what about the 90% of the builders who will of course see such behavior coming soon?

I’m curious what others thing about that and where i’m wrong :slight_smile:

I share the concerns expressed about the GPT Store and its costs. As an early adopter, I see the value of a subscription for professional use, but understand that many will prefer free versions.

The value of specialised GPTs, such as GPT Financial Advisors, lies significantly in the unique skills of their creators. In my case, my 20 years of experience in the field of autism allowed me to develop a specialised GPT as a Social Skill Tutor for people with autism. This example shows how deep knowledge in a specific field can be transformed into a highly effective and specialised GPT, capable of offering a unique service that goes beyond the general capabilities of standard models. Despite the potential of these GPTs, however, I remain sceptical about their ability to attract a large audience willing to invest $20 per month. There are no bikini models old friends to track down or kittens typing on the PC

Regarding the GPT Store, I fear that it could become a closed environment, with users alternating the roles of creators and consumers.

Regarding your first question, I believe that by discontinuing your subscription, your GPTs become unusable, but can be reactivated with a renewal.


GPTs are autonomous agents, and they have a significantly higher computation load possible than just ChatGPT plus itself. They are a replacement for plugins, and even have their own enhanced version of plugins.

A GPT “gizmo” agent, for example, may be provided with a large collection of documents it might browse with a sequence of language model calls. It may make several calls to an external service to build a chain of knowledge or trigger external events on behalf of a user to affect the outside world.

There is no “GPT store without ChatGPT plus”. OpenAI even has their own GPTs that serve more basic purposes, such as with only web browsing enabled, or with DALL-E 3 image creation enabled as a specialization without distraction.

So therefore there is no offering GPT store without offering a full and even more resplendent set of offerings than was previously available in Plus. The desire to have access as seen in this topic (with a misguided plea that they be offered at lower cost instead of the opposite) is a result of their enticement.

1 Like

It seems like the main argument here is “I don’t use it enough to justify $20, so they should make it cheaper.”

Would you make the same argument when buying a new car in an attempt to convince the dealership to lower their price?

I agree with @elmstedt, $20 is a steal.

In a business setting if it saves you 2 hours of work that month you made your money back and then some. If it doesn’t save you two hours you aren’t using it right.

If it still isn’t worth it to you LM Studio is free, go find a good open source model and use that instead.

1 Like

I agree and fortunately even if i run LLM at home for comparisons and ethical assessments most of the time against all APIs out there the custom I tailored for my needs with plus is still unbeatable.

But how long it will unbeatable?

Hopefully builders, users and custom GPTs will be unbeatable for a while, the time needed to make them really unbeatable.

Of course a future model update can make them so far unbeatable.

Said that the unique feature i’m really missing on the new GPT thing is notification sent by app from custom GPTs. Just that feature can open an incredible plethora of enhancements and increase personal use for lot of needs.

Yes I know that can be managed using APIs in some way maybe but the strength is that anyone can tailor a GPT for the mom to remember her to take pills at specific hours and with that feature there’s no need to be a developer to achieve such easy but same time important mission (just an example).

You call it a steal, then, as I have in my previous posts, list a FREE alternative.

Most people don’t use AI. Not yet. And by the time they do, there will be hundreds of alternatives, many with FREE tiers, or at least trial periods.

The GPT Store can be a seed that grows an ecosystem, and draws in new users, or something resembling a product that Google creates; an interesting idea that runs its course, then dies a quiet death.

Nobody that isn’t an AI nerd is going to pay $240 a year for a novelty that they don’t quite know what to do with.

You know how to use it. You have incorporated it into your workflow. But most people are only going to plunk down $ for a refined AI product such as Midjourney ($10 a month), or ($20 a month) than a less refined $20 a month, user-prompted AI.

But if you make the tier more appealing, like $10 a month they might stay subscribed to

  1. keep their custom GPT active

  2. discover new ways to use AI/GPTs.

I agree with trenton.dambrowitz and trenton.dambrowitz that Gpt plus is currently worth its price but, as I wrote earlier, we need to get more people to use Gpts or we will be left ‘preaching to the converted’.

One idea could be to create tokens that can be bought and used occasionally when needed.

1 Like