OpenAI has lifted the five-hour restriction on the use of GPT-5.6 Sol and Codex models for Plus, Pro, and Business subscribers. This decision was made against the backdrop of a sharp increase in load and reaching the milestone of 6 million active users of the company's services.
What Happened
Due to a critical surge in load, OpenAI has temporarily eliminated the five-hour restriction window for paid user categories when working with GPT-5.6 Sol and Codex models. Simultaneously, an update was implemented to increase the computational efficiency of GPT-5.6 Sol, allowing tasks to be performed with lower limit consumption.
Context
The total number of active users of OpenAI services has reached 6 million people. Scaling infrastructure to meet such demand requires not only capacity management but also algorithmic optimization, making increased model efficiency an economic necessity.
Why It Matters for the Industry
For the industry, these changes signal a shift toward infrastructure scaling strategies while maintaining high request density. Optimizing token consumption directly impacts the profitability of cloud computing and could set a new industry standard, where performance evaluation shifts from pure accuracy to efficiency per watt or dollar.
Why It Matters for Users
Developers and professionals now have the ability to work with large-scale code and complex tasks in intensive modes without the risk of quickly exhausting short-term limits. The GPT-5.6 Sol update makes subscriptions more efficient, allowing for faster development cycles (DevCycle) through longer iterative sessions.
What Is Not Yet Known / Limitations
The technical details of the optimization implementation have not been disclosed, making it impossible to determine definitively whether this is an architectural breakthrough or marketing positioning of efficiency. Experts also note the risk of transitioning to a less predictable resource consumption model.
Sources
Author
Look at AI, Editorial Team