OpenAI temporarily eased usage restrictions for GPT-5.6 Sol after demand for the model surged during its first days of broad availability.
Key Points:
- Plus, Pro and Business users temporarily received relief from the rolling five-hour usage restriction.
- OpenAI also reset current usage allowances, giving affected users immediate access to additional capacity.
- The company said efficiency changes should make GPT-5.6 Sol consume less of each user’s available quota.
GPT-5.6 Limits
The company confirmed that it was temporarily removing the five-hour restriction for Plus, Pro and Business plans after heavy activity across Codex and ChatGPT Work during the previous 48 hours.
OpenAI product lead Tibo Sottiaux announced the change on X, saying the company would also reset existing usage for all affected customers.
Codex and ChatGPT Work count local messages and cloud tasks against a shared allowance. Under the usual system, users can reach a rolling five-hour cap, while separate weekly restrictions may also apply according to the plan and selected model.
Removing the shorter window gives paid users more room to continue coding, research and agent-based tasks without waiting for the five-hour allowance to renew. The change remains temporary, and OpenAI has not described GPT-5.6 Sol as offering unlimited access.
Also Read: Byreal Clears $3.7B In Volume In Its First Year On Solana
OpenAI Efficiency
Sottiaux said OpenAI was also rolling out changes intended to make GPT-5.6 Sol more efficient across its products. The company expects those adjustments to reduce how much usage each task consumes, allowing customers to complete more work before reaching their limits.
OpenAI did not provide a technical explanation for the improvement. Lower token consumption could be one factor, but the company has not confirmed the mechanism, so any specific explanation remains speculative.
The one-time reset immediately restored additional capacity for users who had already consumed much of their allowance.
It also addressed complaints from customers who reported reaching the five-hour threshold quickly while using the model for demanding coding and agentic workloads.
GPT-5.6 Sol launched as OpenAI’s most capable model for complex professional work, with access spanning ChatGPT, ChatGPT Work, Codex and the API. Its early usage pressure shows how quickly new model releases can test capacity controls, particularly when local interactions and longer cloud tasks draw from the same quota.
Read Next: Bitget Launches Cash Plus To Pay Yield On Idle Stablecoin Balances