Written By Deepti Ratnam
Written By Deepti Ratnam
Edited by Deepti Ratnam |Published: Aug 07, 2026, 09:37 AM (IST)
OpenAI removes text chat limits for ChatGPT Free users
ChatGPT Free users are getting one of the biggest upgrades. OpenAI has announced that it is lifting the restrictions on chat-based conversations that rely solely on text. The AI tech giant is lifting of restrictions on text-based chat, so that people can become more free to continue their conversations without hitting the message limit.
The announcement follows shortly after ChatGPT became the first AI app to reach 1 billion users a week. In addition to free text chats, OpenAI will be rolling out new AI models and features for Free, Go and Plus subscribers. While the changes are different for each plan, the goal is the same and which is to make ChatGPT more capable, accurate and flexible.
The most notable difference is for users with the Free plan. According to OpenAI, text conversations will no longer be limited, which will make ChatGPT more convenient to use throughout the day without being interrupted.
This same benefit is also being extended to Go users.
However, the change only applies to text chats. Other features will continue to have separate usage limits. These include file uploads, image uploads, voice conversations and AI image generation.
This will allow users who primarily use ChatGPT for writing, asking questions, brainstorming ideas or learning something new to continue chatting without a text message limit.
5.6 Sol much better in chat now
and unlimited text chat for free users! https://t.co/zgqGqI9qfR
— Sam Altman (@sama) August 6, 2026
In addition to open chats, Free and Go users will also receive a new default AI model.
OpenAI has upgraded its GPT-5.5 model to GPT-5.6 Luna. After the rollout, users will not need to manually change models as Luna will do the rest for regular conversations.
The company claims that GPT-5.6 Luna offers more consistent results and maintains a natural and fast conversation. It will be able to do better in its ability to facilitate the user in routine activities, from answering straightforward questions to explaining or helping with writing.
The new Think button has been added to the Free and Go.
The feature is intended for questions that call for more reasoning than a typical conversation. ChatGPT takes more time to solve the problem before providing a solution, as opposed to reacting instantly.
This may be helpful for problems with multiple steps, learning about complex concepts, or organizing a project or answering questions that require more detailed analysis.
Whenever users want ChatGPT to think more before answering, they can click on the Think option.
Premium subscribers are also getting new features.
ChatGPT Plus and Pro users are getting an upgraded version of GPT-5.6 Sol. The company claims that the model works better when it comes to daily productivity such as answering questions, conducting web research, making project plans, writing content, providing advice and making decisions.
The new edition also aims to yield more concise and unambiguous responses without reducing accuracy.
OpenAI explained that the change in the model was for the GPT-5.6 Sol, not Codex and Work, and that the latter is not changing.
Plus and Pro customers will also be receiving a thinking slider with this new model.
The slider enables users to determine the level of reasoning that ChatGPT will apply before giving a response. If the question is simple, they can keep the depth of the thinking level down and are able to get a quicker answer. They can up the ante if it’s a more complex request, helping to raise the thinking level, and extending the model’s time available for analysis before he or she provides the answer.
The feature allows the user to have more control on the level of response speed and depth of reasoning.
In addition, OpenAI has released their internal evaluation metrics of their newest models.
The company claims that GPT-5.6 Luna also makes 62 percent fewer factual mistakes than GPT-5.5 Instant. GPT-5.6 Sol, on the other hand, cuts down on factual errors by 68 percent from the previous model.
These numbers are based on OpenAI’s own tests, but they indicate that the company is working to improve the reliability of ChatGPT in its everyday application, particularly in scenarios where precision counts.
The new GPT-5.6 Sol model is already in use for ChatGPT Plus and Pro users.
The rest of the updates for free and go users will start rolling out next week. One of the biggest changes in the free version of ChatGPT, AI experts say, will be the new Think button, which will offer unlimited text chats, in addition to the enhanced quality of the tool’s performance and the ability to access thousands of additional web pages.
