OpenAI has quietly dismantled one of ChatGPT’s most persistent barriers to entry, removing all limits on text-based conversations for free users as the service crosses 1 billion weekly active users.
The change, announced August 6, 2026, represents a fundamental shift in how the company positions its flagship product. Free-tier users—who previously faced message caps that frequently interrupted workflows—can now engage in unlimited text conversations powered by the new GPT-5.6 Luna model. The updated GPT-5.6 Sol model, meanwhile, arrives for Plus and Pro subscribers with more compact answers and adjustable reasoning controls.
The move comes at a pivotal moment. ChatGPT’s user base has grown to the point where limiting free access arguably cost OpenAI more in friction than it saved in compute expenses. With 1 billion weekly users, the service has become so integral to daily digital life that capping conversations increasingly frustrated the very audience OpenAI needs to convert into paying customers.
A More Capable Free Tier
The new GPT-5.6 Luna model replaces GPT-5.5 as the default for Free and Go users. OpenAI’s internal evaluations suggest a meaningful jump in accuracy, with factual errors occurring 62% less frequently compared to the previous instant-response model.
Free users will also gain access to a “Think” button, allowing them to select higher reasoning power for complex questions. This is a notable departure from the traditional binary approach to AI model access, where advanced reasoning capabilities were either entirely gated behind paywalls or hidden behind complicated prompt engineering.
The thinking controls for paid users are even more granular. Plus and Pro subscribers receive a thinking slider that adjusts how much cognitive effort the model applies to each query. This represents a more sophisticated approach to AI interaction than the current “one model fits all” paradigm that dominates most consumer AI products.
The Two-Tier Model Strategy
OpenAI has effectively created a dual-purpose system with GPT-5.6. Luna handles the high-volume, low-complexity interactions that characterize most free-tier usage. Sol, available to paying subscribers, focuses on delivering robust, compact answers for quick tasks including web research, planning, writing, and decision-making.
This bifurcation reveals something interesting about OpenAI’s product strategy. Rather than simply providing a more capable model to paying users, the company is optimizing each model for different use cases. Luna prioritizes cost efficiency and broad accessibility. Sol prioritizes answer quality and speed for users who depend on ChatGPT for professional work.
The new “thinking slider” for Plus and Pro users is particularly noteworthy. Users can now tune the model’s reasoning depth based on query complexity, a feature that acknowledges different tasks demand different levels of cognitive investment. A simple fact-check doesn’t require the same processing as a complex business strategy analysis.
What This Means for Users
For free users, the practical impact is immediate. No more waiting for message caps to reset. No more carefully rationing queries. This changes the calculus for millions of users who rely on ChatGPT for everything from research to content creation.
The improved accuracy of GPT-5.6 Luna also addresses one of ChatGPT’s persistent pain points. A 62% reduction in factual errors represents meaningful progress, though OpenAI’s internal evaluation methodology remains undisclosed. Users should still verify critical information from primary sources.
The Think button gives free users access to capabilities that previously required paid subscriptions. While the feature likely has usage limits—OpenAI has confirmed separate limits remain for files, images, voice, and image generation—it dramatically expands what free users can accomplish.
The Competitive Landscape
Google’s Gemini and Anthropic’s Claude both offer free tiers with varying limitations. OpenAI’s decision to remove text-based limits entirely puts pressure on competitors to follow suit. If unlimited text chat becomes the industry baseline, the competitive differentiator shifts to other features: multimodal capabilities, reasoning depth, customization options, and ecosystem integration.
The timing also matters. ChatGPT hitting 1 billion weekly users makes the service a platform rather than just a product. At this scale, OpenAI can afford to treat text-based interactions as a loss leader, using free access to drive adoption of paid features, enterprise services, and developer tools.
OpenAI’s Changing Business Model
Here’s what the announcement doesn’t explicitly state but is worth examining: OpenAI is betting that unlimited access to basic text interactions will drive more paid upgrades than the previous limitation model.
The company has effectively concluded that the conversion funnel works better when users experience ChatGPT without friction. If you can use the service freely for routine tasks, you’re more likely to pay for enhanced capabilities when you need them. This is fundamentally different from the approach where free users constantly hit walls, potentially driving them to explore competitors with more generous limits.
The new thinking slider for paid users suggests OpenAI recognizes that the next battleground is not access but control. Users want to direct how much reasoning power their AI assistant applies, not just unlock better models. This mirrors trends across software, where power users increasingly demand granular control over performance characteristics.
What Comes Next
Next week, free users will gain access to unlimited text chats and the Think button. The upgraded GPT-5.6 Sol model is available for Plus and Pro users today.
The question now is whether OpenAI will extend the unlimited approach to other modalities. Images, voice, and file uploads still carry limits. If text-based access becomes truly unlimited, users will likely demand the same treatment for other input types.
OpenAI has also set expectations for continued rapid iteration. The company mentioned that GPT-5.6 Sol for Codex and Work remains unchanged, suggesting future updates will bring the enhanced capabilities to those specialized versions as well.
For users, the immediate takeaway is simple: ChatGPT’s free tier just became dramatically more useful. The long-term implications, however, extend far beyond message limits. OpenAI is signalling that the era of artificial scarcity in consumer AI is ending, replaced by a more sophisticated model where value extraction happens through depth rather than breadth of access.
The real test will be whether this approach proves more sustainable. If unlimited free text access drives enough paid conversions, expect Google and Anthropic to follow. If not, OpenAI may need to reconsider its strategy. But with 1 billion weekly users and improving models, the company is betting that the path to profitability runs through removing barriers rather than building them.
