SAN FRANCISCO, CALIFORNIA / RankWire.AI / – OpenAI is broadening access for Free and Go users to engage in unlimited text conversations powered by GPT-5.6 Luna. The company announced that this update will commence next week and will be maintained with safeguards to prevent misuse. During the current rollout, GPT-5.6 Luna becomes the default model for these users. Additionally, free users will see a new Think button that grants the model extra time to process complex questions before delivering a response.
This unlimited text access specifically pertains to interactions within ChatGPT. Limits for file uploads, image creation, and other specialized tools that require additional processing remain in effect. The Think feature will launch alongside the expanded text quota next week, enabling Luna to take more time to reason through questions before generating an answer. Consequently, ChatGPT users on Free and Go plans will enjoy broader access to the core chat functionalities, although not all features will be available with unlimited use.
GPT-5.6 incorporates Luna, Sol, and Terra, each designed to meet different levels of performance and computational needs. Luna is positioned as the faster and more cost-effective choice within this family. ChatGPT now reportedly reaches around one billion users weekly across activities such as search, research, planning, and writing. The current rollout places Luna as the default experience for Free and Go accounts, while distinct rules remain for tools outside standard text conversations.
Enhanced GPT-5.6 Sol controls now available for paid subscribers
Subscribers to ChatGPT Plus and Pro are now receiving an upgraded version of GPT-5.6 Sol for their daily chats. Sol supports both quick replies and in-depth reasoning within paid plans. The company stated that the updated model offers more focused responses and minimizes unnecessary details when a shorter reply suffices. Paid users also gain access to a reasoning slider across web, mobile, and desktop platforms, allowing them to control how much processing ChatGPT applies to tasks like research, planning, writing, and coding.
Internal testing indicates that the newest models exhibit reduced factual error rates. The assessments involved prompts related to finance, healthcare, and legal topics demanding detailed factual answers. Results showed approximately 62% fewer responses containing at least one factual mistake for GPT-5.6 Luna compared to GPT-5.5 Instant. Similarly, GPT-5.6 Sol demonstrated a reduction of about 68% in errors under the same testing conditions. The company also highlighted improved accuracy in handling dates, numbers, sources, rules, and assumptions when these details impacted the responses.
Safety measures remain in place across the expanded ChatGPT platform
The latest models feature additional safeguards for users identified as under 18. The safeguards address areas such as romantic roleplay, age-restricted challenges, and claims that might substitute real-world relationships. They also set boundaries concerning sexual content, eating disorders, dangerous activities, restricted products, and graphic violence. The company mentioned that new evaluations have been introduced to monitor responses involving younger users. Unlimited text interactions for Free and Go accounts will continue to operate within the platform’s existing abuse prevention protocols.
This update mainly concerns the core ChatGPT experience and does not alter every product that utilizes GPT-5.6. The GPT-5.6 Sol versions used in ChatGPT Work and Codex remain unaffected by this release. Paid subscribers can now access the updated Sol model and reasoning controls. Meanwhile, Free and Go users will see Luna as the default model starting this week. The unlimited text chats and Think button will be available next week, while uploads, images, and other tools will continue to have separate usage restrictions.