OpenAI cancels text chat restrictions on ChatGPT, GPT-5

OpenAI cancels the text chat restrictions of ChatGPT, and the GPT-5.6 series models are fully upgraded. OpenAI recently announced that it will cancel the text chat restrictions of ChatGPT for all users, further expanding the scope of use of AI assistants. Currently, the number of ChatGPT weekly users has exceeded 1 billion, and the new round of model upgrades will be supported by the new GPT-5.6 series. Among them, GPT-5.6Luna will become the default model for Free and Go users, replacing the previous GPT-5.5. In the future, users will also get a new "Think" button, which can actively select a higher reasoning mode to deal with more complex problems. OpenAI stated that capabilities such as file processing, image input, voice functions, and image generation will still maintain independent quota limits. For Plus and Pro users, OpenAI simultaneously launches the upgraded version GPT-5.6Sol model. This model is mainly designed for high-frequency fast tasks, including daily Q&A, web search, suggestion provision, planning, writing and decision support. Officials say it can provide more concise and reliable answers. It should be noted that the GPT-5.6Sol version used in Codex and Work products has not been adjusted this time. Additionally, Plus and Pro users will have access to a “think level” slider that allows them to adjust the amount of computation put into their models based on task complexity and required reasoning steps, balancing speed of response with depth of answers. The internal test data released by OpenAI shows that compared with GPT-5.5-Instant, the fact error rate of GPT-5.6Luna is reduced by 62%, and the fact error rate of GPT-5.6Sol is reduced by 68%. Currently, GPT-5.6Sol is open to Plus and Pro users. Related updates for Free and Go users will be launched this week. It is expected that all users will be able to experience unlimited text chat and the new "Thinking" function starting next week. This upgrade further reflects OpenAI’s direction of promoting the development of AI assistants from simple question and answer tools to long-term, high-frequency intelligent interaction platforms. As the scale of users continues to expand, model reliability, reasoning capabilities and usage quota optimization are becoming important factors in the competition of AI products. via AI News (author: AI Base)