OpenAI Removes Text Limits for Free Users in GPT-5.6 Model Upgrade
The company is deploying GPT-5.6 Luna and Sol models to expand accessibility and introduce granular reasoning controls.
OpenAI has updated ChatGPT to provide free and "Go" tier users with unlimited text chats and enhanced reasoning capabilities. The move signals a strategic push to lower the barrier to entry for millions of users while refining the precision of its latest AI models.
Free and Go users now have GPT-5.6 Luna as their default model, which includes a new "Think" button designed for deeper reasoning. Meanwhile, Plus and Pro subscribers have been upgraded to a more reliable version of GPT-5.6 Sol. This premium version introduces a reasoning slider, allowing users to manually adjust the level of "thought" or computational effort the model applies to a specific response.
According to internal evaluations released by OpenAI, these updates significantly reduce hallucinations. Factual errors were 62% less common with GPT-5.6 Luna and 68% less common with GPT-5.6 Sol when compared to the previous GPT-5.5 Instant model. Despite the unlimited text capabilities for free users, OpenAI confirmed that limits remain in place for file uploads, image generation, and other specialized tools.
Expanding the Intelligence Gap
This rollout follows a broader industry trend of expanding high-tier model access to non-paying users to drive growth and market share. By removing daily text limits, OpenAI is positioning ChatGPT as a more ubiquitous utility, potentially squeezing competitors who maintain stricter paywalls for their most capable models.
Furthermore, the introduction of the "Think" button and the reasoning slider represents a shift in user experience. Rather than the AI deciding how much processing power to use behind the scenes, OpenAI is giving users explicit control to balance speed against accuracy. This allows professionals to opt for high-effort reasoning for complex tasks while maintaining rapid responses for simple queries.
Future Outlook
OpenAI described the update as a "concrete step toward more abundant intelligence," emphasizing the goal of making reliable answers more widely available. However, the rollout is not universal across all products; the company noted that the version of GPT-5.6 Sol powering ChatGPT Work and Codex remains unchanged in this release.
Industry observers will now be watching to see if this increase in free-tier accessibility impacts Plus and Pro subscription retention, or if the new granular controls provide enough unique value to keep power users paying. For now, the focus remains on the transition from the GPT-5.5 era to the more stable, reasoning-heavy 5.6 architecture.