OpenAI is testing a 3,000-per-week limit for GPT-5 Thinking mode, aiming to balance advanced reasoning features with system performance. The cap applies to ChatGPT Plus users and marks a shift in how the company manages access to its most resource-intensive AI capabilities.
Why the Limit Exists
The Thinking mode in GPT-5 is designed for deeper, more structured reasoning. It consumes significantly more computing power than standard GPT-5 interactions. By introducing a 3,000-per-week limit, OpenAI hopes to ensure stable access for all subscribers while continuing to refine the feature.
User Trends Driving the Change
Since GPT-5’s launch, usage of Thinking mode has grown rapidly. For free users, it jumped from less than 1% to around 7% of daily activity. Among Plus subscribers, it surged from 7% to roughly 24%. This spike has placed higher demand on infrastructure, prompting OpenAI to manage capacity more carefully.
Sam Altman’s Plans for Expansion
OpenAI CEO Sam Altman confirmed that the company intends to raise rate limits across all GPT-5 model variants in the future. This includes the main, mini, and Thinking models. Altman’s comments suggest that the 3,000-per-week threshold is temporary and could increase as OpenAI scales server capacity.
Interface Updates to Reduce Confusion
Many users have noted difficulty in identifying which GPT-5 model is active during a session. OpenAI plans to roll out a visual indicator within the ChatGPT interface to make this clear. This should help users understand when they are in Thinking mode versus the main or mini versions of GPT-5.
Conclusion
The GPT-5 Thinking mode limit is a strategic move to maintain quality and performance while demand grows. With plans to expand access and improve user clarity, OpenAI appears focused on delivering more powerful reasoning tools without compromising stability. As capacity increases, users can expect the cap to be lifted or expanded.
0 responses to “GPT-5 Thinking Mode Gets 3,000-Per-Week Limit in OpenAI Tests”