OpenAI Expands ChatGPT Free Tier: Unlimited Text Chats and GPT-5.6 Luna for Free Users

What happened

OpenAI announced a significant expansion of its ChatGPT free tier, set to take effect next week following the August 6, 2026 announcement. The changes include unlimited text chats for both free and Go tier users, removing the previous rate limits that applied to text-based conversations. Additionally, OpenAI is introducing a new "Think" button feature for free and Go users, enabling access to higher reasoning capabilities when tackling more complex questions.

The model landscape also sees an upgrade: ChatGPT's default AI model for free users is being upgraded to GPT-5.6 Luna, which was recently released. Meanwhile, Plus and Pro subscribers will receive GPT-5.6 Sol, a variant featuring improvements in factual reliability, fewer mistakes when answers depend on dates/numbers/sources/rules/assumptions, and better utilization of sources found while answering questions.

For Plus/Pro subscribers, OpenAI is also rolling out a new slider that allows users to decide "how much thought ChatGPT puts into an answer." The updated GPT-5.6 Sol model and the thinking slider will be available starting Thursday following the August 6, 2026 announcement date.

OpenAI describes the new GPT-5.6 Sol as providing "more direct responses" with "tighter formatting," and ChatGPT will "avoid extra detail when it does not help" for Plus/Pro users.

Importantly, file uploads and images will continue to have rate limits even after these changes take effect, according to the company's announcement.

Why it matters

This expansion signals OpenAI's strategic pivot toward broadening its user base while maintaining premium differentiation. By removing text chat rate limits for free and Go users, OpenAI is effectively making ChatGPT more accessible as a general-purpose AI assistant without requiring a paid subscription. This move could significantly increase daily active users on the free tier, potentially creating a larger funnel for premium conversions.

The introduction of GPT-5.6 Luna as the default model for free users represents a substantial upgrade in capability distribution across tiers. Previously, advanced models were largely gated behind paid subscriptions; this change democratizes access to cutting-edge AI reasoning at no cost. For free and Go users, this means they can now tackle complex tasks—research, coding assistance, creative writing, and more—without hitting artificial conversation limits that previously constrained productivity.

The "Think" button feature adds another layer of sophistication, allowing free users to opt into deeper reasoning modes when needed. This is particularly valuable for students, researchers, and developers who need the AI to work through multi-step problems or complex code generation tasks.

For Plus/Pro subscribers, the new slider provides unprecedented control over the trade-off between response speed and depth of reasoning. Users can now customize their experience based on whether they need quick answers or want ChatGPT to "think harder" about nuanced questions. The improvements in factual reliability and source utilization for GPT-5.6 Sol address common complaints about AI hallucinations, making it more trustworthy for professional use cases.

The announcement also reinforces OpenAI's tiered strategy: free users get powerful tools with unlimited text chats and the latest model (GPT-5.6 Luna), while Plus/Pro subscribers receive enhanced reliability features, source-aware responses, and fine-grained control over reasoning depth. This maintains clear value differentiation while expanding the overall user base.

What to watch

Several developments warrant close attention in the coming weeks:

1. User adoption metrics: How quickly do free users adopt the unlimited text chat feature? Will they increase their usage significantly, or does the previous rate limit already constrain most users?

2. Premium conversion rates: With such generous free tier offerings, will OpenAI see increased conversions to Plus/Pro as users experience GPT-5.6 Luna's capabilities firsthand? Or will the expanded free tier reduce conversion pressure?

3. Competitor responses: How do other AI companies (Anthropic, Google, Microsoft) respond to this expansion? Will they match these features or differentiate elsewhere?

4. Model performance in production: Real-world usage of GPT-5.6 Luna on the free tier will reveal how it performs at scale compared to GPT-5.6 Sol. Users and developers will be watching for any degradation in quality when serving many more requests.

5. The thinking slider adoption: Will Plus/Pro users embrace the new control mechanism, or do they prefer OpenAI's default behavior? This could inform future AI UX design across the industry.

By the numbers

Source snapshot

Sources: https://www.theverge.com/ai-artificial-intelligence, https://www.theverge.com/ai-artificial-intelligence/976239/openai-chatgpt-free-go-text-chats