OpenAI eases GPT-5.6 usage limits, learning from Gemini's restriction troubles

OpenAI is learning from Gemini's trouble with usage limits and easing restrictions on GPT-5.6

The team is also working on optimizing usage with ChatGPT Work and multi-agent workflows. As with every major feature release from OpenAI, the new GPT-5.6 models have already been overused beyond expectation. The models are officially reported to have reached 6 million active ChatGPT users, but instead of crippling them with limits as Google did, OpenAI is rewarding them with faster usage resets. In a post on X, Thibault Sottiaux (Tibo), OpenAI's newly appointed head of core products, announced the temporary removal of the 5-hour usage limit restriction, but only for Plus, Business, and Pro users. That should mean you can burn through your entire week's limit if you want to. There's no comment on validity, but with Anthropic extending Fable 5 access for another week, OpenAI may be feeling obligated to one-up it as well. In a separate post, Tibo also addressed concerns about excessive usage and said the team is experimenting with different reasoning efforts (or juice values), especially in multi-agent scenarios. While the reasoning capability was briefly reduced as part of a test, it has been reverted to its original value. OpenAI recently launched its GPT-5.6 family of AI models after a short preview window. These models come in three different forms: Sol, Terra, and Luna, with varying performance and compute requirements. Of the trio, GPT‑5.6 Sol is the most formidable and is intended for demanding chores, such as coding, creative design, and deeper analyses. Luna is designed for cost-efficient tasks, despite which it is claimed to outperform Anthropic's Claude Opus 4.8 and 5. Meanwhile, Terra offers a balance between the two. Along with announcing these models, OpenAI has also released its answer to Claude Cowork. It's called ChatGPT Work, and it works similarly by gathering context from across apps and services on your desktop or in a web browser and coordinating the tasks required to achieve your professional goals for you (even though there's nothing stopping it from helping with your personal ones, either). Along with this announcement, OpenAI has also deprecated Codex, essentially merging it with Work. These new abilities have led to record-breaking usage among ChatGPT users, which is why today's announcement sounds welcome.

Want to try the latest ChatGPT and Claude models? Now's your chance

This competitive move allows users to freely explore advanced AI reasoning capabilities and multi-agent features without typical restrictions. If you haven't kicked the tires yet on the latest and greatest ChatGPT and Claude models, this is your lucky week. OpenAI is (fittingly) opening the flood gates to GPT-5.6 Sol, its just-released and most powerful model, announcing Sunday that it's "temporarily" lifting the five-hour usage window for ChatGPT Plus, Business, and Pro subscribers. At the same time, Anthropic is - again - extending the trial period for Fable 5, its own new top-of-the-line, giving Claude subscribers another week of in-plan access. Anthropic is also doubling Claude Cowork usage limits until August 5th and boosting weekly Claude Code limits by 50 percent through July 19th. The relaxed usage limits are only temporary, of course, and the cynical way of looking at them is that Anthropic and OpenAI are just whetting our appetites before tighter usage restrictions go back into effect. Then again, it could simply be the result of competition between the rival AI giants, with OpenAI trying to scoop up Anthropic subscribers annoyed by the previously announced and tighter restrictions on Fable 5, and Anthropic responding in kind. Whatever the reason, now is a good time for ChatGPT and Claude subscribers to reap the benefits of the relaxed usage limits, particularly when it comes to GPT-5.6 and Fable. Aside from (temporarily) lifting its usual five-hour usage window, OpenAI says it's optimizing the GPT-5.6 Sol's reasoning and "mutli-agent" use to keeping it from burning so many AI tokens. For now, it appears GPT-5.6 Sol will remain on ChatGPT subscription plans, with no indication from OpenAI that subscribers will need to buy usage credits for the new model (aside from those who've burned through their regular usage limits). Anthropic has warned it will eventually pull Fable 5 from its subscription plans, although an Anthropic exec has (now) repeated an earlier promise that "we aim to restore Fable as a standard part of our subscriptions as soon as capacity allows."