The OpenAI 200 Knife Pro package resumes, but the limit is low!

  It's the first time in the history of the world.

OpenAI's $200 Pro subscription service is finally announced to reopen tomorrow.

Sounds like a good thing.

Then, "Godfather" Tibo broke the rules for calculating the amount of ice-cream in a large basin, causing the efficiency to plummet by 50%!

In the eyes of DevDay, the heavy king has been blown up, but the calculator configuration is in advance.

While Tibo posted a long post in advance to try to calm the mood, the comments quickly turned around. Global developers burned in anger and shouted, "OpenAI, are you making us a mess?"

It looks like this is the transition from subscription systems to full API algorithms.

In the GPT-6 era, are we still free to calculate?

A few hours ago, Tibo made an unexpected announcement in advance of a major update.

"Well, tomorrow we'll reopen the Pro subscription for new users at $200 a month".

When the phrase came out, the eyes of many developers who were waiting for breastfeeding due to insufficient computing power were lit.

It's worth noting that the $200 Pro subscription - the previous 20x plan - has always been the lifeblood of AI developers and startup teams.

But right after that, Tibo's words changed:

However, what we did in sync with the restart was change the way we calculate consumption.

If you look closely at the bill, you'll find that the equivalent amount of API consumption included in the new version of the subscription is half that of the old $200 Pro plan.

A stone raises a thousand waves!

Faced with the shrinking statement, Tibo frankly blurted out a "I'm good for you" gesture, giving four explanations:

1.No cooling restrictionsOfficial pledge not to restart "five-hour cooling" restrictions.

2.Painting a Long BreadModel efficiency and throughput will continue to improve, and efficiency dividends will be returned through APIs.

3.Rejection of low pricesThis week, the GPT-6 Sol and GPT-6 Luna API have dropped 50 percent, with the goal of gradually bridging the gap between on-demand and subscription prices.

4.Bury the DevDay FountainTomorrow Pro subscriptions will receive all new rights, and these huge features are completely excluded from the daily allowance.

At the end of the post, he said: "I want to be transparent and share this change in advance before the good news of tomorrow's earthquake comes to you"...

The answer is yes, of course not!

Tibo's pulmonary embolism not only does not act as a fire extinguisher, but rather as an oil burner.

The comments area was instantly overwhelmed with traffic jams, developers and angry users.

The central point of everyone's anger is very concentrated: What are you doing with the big pie of the future to take away my real gold and silver?

The most powerful of these is the well-known developer Lukas Böhm.

He directly tore apart Tibo's logical defensive line, which he called a "counterattack" at the level of a textbook:

"You can't start by saying, 'Why can't you do more work now than you did a month ago,' and then come up with an excuse that's nonsense!

(a) The permanent abolition of the five-hour limit does not help us to do 'more' work, only replaces the time;
(b) How can the promise of 'increasing efficiency in the future' be used to bridge the gap between 'tomorrow' and 'a month ago'?
c) Do you adjust the pricing incentive to how much work we can do that has a half-paying relationship to it?
d) Increasing the so-called 'non-consuming ecosystem functionality' doesn't mean we can do more on actual Token output!"

In the end, Lukas leaves behind an even more disturbing statement: "Please don't become a PR nightmare like Anthropic, which is why so many people leave and come here looking for a better future!"

This is the first time I've ever been in a relationship with a woman.

Brother, I'm a developer, I don't want any candy, I just want high-quality code, lots of fast code!

The great V Ashutosh Shrivastava even wrote a note:

"OpenAI, if I'm not mistaken, is about cutting the equivalent use of 200 knives in half and allowing you to believe that future efficiency and price reductions will bridge the gap".

Your 20-fold package has now become 10-fold; your 5-fold package has become 2.5 times.

Finally, he joked: "Happy OpenAI Day to all developers!"

Some netizens have even directly moved the strongest bids out of the box:

「Don't write such a long post to get dirty. As long as you can beat Claude Opus 5.5 on The Whole Value of a Dollar, what do you think?」

In the face of this incredibly harsh comment, Tibo replied: "Agreed".

Why would OpenAI, which has always paid great attention to the developer ecosystem, choose to make such a defeatist decision at such a critical juncture?

Is it really just for "transparency"? Combined with the dynamics of the recent GPT-6 series of models, this may be a well-thought-out "conspiracy" by OpenAI.

1. The ultimate bet on subscription versus paid-for-API

In the past, many heavy-duty users bought the 200-knife Pro (i.e. the 20x plan) because it was much more cost-effective than directly charging the API. In fact, it was officially subsidized with "loss" calculations in exchange for advanced user retention and eco-stickness.

But with the GPT-6 Sol and GPT-6 Luna both models falling directly by 50% this week, the cost structure of OpenAI has changed.

If the API is cheap enough and subscriptions to 200 knives are still as efficient as they were, then these Pro users will be a huge black hole for OpenAI.

Tibo's logic is this: I've reduced the single price by 50% so the number of times you buy the same money doesn't change, but the "total price" is halved.

But the developer's logic is this: Since you're 50% off, I'm using the same 200 knives, and you should give me twice as many calls!

This is the irreconcilable contradiction between the two parties. OpenAI no longer wants to be the boss, they try to drive large customers gradually from "packaged luncheon" to a model of on-demand billing.

2. "Class Barriers" of the GPT-6 era

What's funny is that netizens @tinkerersanky pulled out a ChatGPT model to disguise the chain.

Plus users: $20/month:The GPT-6 Luna is the only basic experience.

Pro 5x users: $100 a month:GPT-6 Luna and some GPT-6 Sol can be used (low/medium intensity).

Pro 20x users: $200 a month:GPT-6 Sol can be used (high/ultrahigh intensity).

Pro 50x users: $500/month: to touch the true peak of the GPT-6 Astra.

The truth is out there.

The scarcity of arithmetic did not decrease with the exponential rise of large model capacities, but instead became more expensive due to the emergence of more advanced models such as Astra.

OpenAI is cutting half of the $200 Pro program, which is essentially in the works.Discrimination against prices and traffickingThey're no longer willing to spend $200 to give you an unlimited taste of the incredibly expensive calculations of the GPT-6 era.

As the netizens mock: "500 dollars will be a new 20-fold plan. Smart is getting cheap and unmeasurable, but top-notch smart is still going to hurt you".

For tomorrow's "Real King Bore" to make room for profits

At the end of the article, Tibo repeatedly stressed: "Tomorrow we're going to be adding a whole new amount of non-consumable subscriptions... there are a lot of exciting new things that will make subscriptions super attractive".

It's a great way to get rid of computing power in order to give way to new functions.

What exactly is this? an infinite context memory pool that doesn't need to consume tokens? a built-in super powerful Agent automation execution framework?

In any case, OpenAI's strategy is clear:We're going to「Salesforce」It's like a box of chopsticks.「Selling integrated AI workflows」It's an ecological masterpiece.

Computing is no longer the core selling point of subscriptions, and AI is the original function.

In this wave, the name is mentioned the most times, except for OpenAI, which is its death against the title.Anthropic。

The netizens use it directly.「Beat Opus 5.5 in the overall value for a dollar」General Tibo (beating Opus 5.5) at a value of every dollar.

This illustrates an extremely cruel state of affairs in the industry:Developers are not loyal. Whose models write more code, whose APIs are cheaper, whose restrictions are less, and who brings money home.

Previously, Anthropic lost a large number of developers due to sealing issues, stringent speed limits and poor public relations. They came to OpenAI with the hope of purchasing the 200-knife Pro service and thought they had found a safe haven.

As a result, OpenAI's "Midnight Thieves" operation allowed the "refugees" to feel the sting again.

If tomorrow's DevDay, OpenAI, or the so-called "mysterious, unstoppable new capabilities" aren't able to bring out enough GPT-6 killer capabilities to suppress the entire game, it's likely that this wave of heavy-duty developers who are "disgusted" will be swept back into the arms of Claude Opus 5.5.

OpenAI is rooting for the new stuff they're releasing tomorrow, enough to make developers forget about the 50% cut today.