OpenAI significantly reduces paid users’ GPT-4 query frequency to 25 times in 3 hours

Upon OpenAI’s initial launch of ChatGPT subscriptions, they likely did not anticipate such a high rate of paid memberships, leading to an overextension of subscription quotas.

When GPT-4 was unveiled this week, subscribers were granted priority access, permitting 100 queries every four hours, which was subsequently reduced to 50 queries.

Presently, subscribers are restricted to a mere 25 queries within a three-hour window, and it appears probable that OpenAI may further adjust subscription quotas next week.

Yesterday, OpenAI experienced a temporary suspension of ChatGPT’s free version access and an API malfunction, while ChatGPT Plus subscribers remained unaffected.

However, the current quota adjustments presumably stem from server overload, and OpenAI has no choice but to limit query frequency to ensure GPT-4’s stability.

Moreover, the quota restrictions suggest an exceptionally high number of subscribers, yet if everyone were to subscribe, it would be tantamount to having no subscribers at all.

As for GPT-3.5, subscribed users face no limitations; even during peak loads, OpenAI merely restricts free version usage.