OpenAI ends ChatGPT free text chat limits
OpenAI is ending ChatGPT free text chat limits, a change announced Thursday that also switches Free and Go users over to GPT-5.6 Luna as the default model. Unlimited text chats and a new Think button follow next week. Plus and Pro subscribers get a retuned GPT-5.6 Sol with a reasoning-depth slider that starts rolling out now.
The model swap happens this week. Luna, introduced in early July and designed for speed, replaces GPT-5.5 Instant as the default on the free tier and OpenAI's low-cost Go plan. Starting next week, text-only prompts stop counting against a rate limit, so free users can keep as many conversations going as they want, with no daily cutoff. The Think button lets Luna spend extra time on harder questions, one message at a time, with abuse guardrails in place.
What removing ChatGPT free text chat limits means for you
OpenAI's internal evaluations measured how often each model returned an answer with at least one factual mistake. Luna was 62% less likely than the previous default to do so, and the reworked Sol was 68% less likely. For Plus and Pro users, Sol was retuned for everyday conversation rather than benchmark-style tasks, a shift that OpenAI says yields answers that stay on topic and facts that hold up more often. A new slider controls reasoning depth per response, letting paid users choose between quicker replies and deeper analysis. Only ChatGPT uses this chat-optimized edition of Sol.
The Think button and the slider do the same job on different tiers: free users get a per-message boost for hard questions, while paid users get continuous control over how much reasoning each answer uses.
What stays limited
The uncapping applies to text only. File uploads, image messages, voice tools and image generation keep separate limits for Free and Go users, and those caps remain subject to abuse safeguards.
| Tier | Default model | Text chat limits | New control |
|---|---|---|---|
| Free / Go | GPT-5.6 Luna (this week) | Unlimited (next week) | Think button |
| Plus / Pro | GPT-5.6 Sol (retuned) | Unchanged | Reasoning-depth slider |
The end of ChatGPT free text chat limits arrives as the platform passes 1 billion weekly users. For years, the free-tier cap was one of the main pressures steering casual users toward a paid subscription. Removing it changes the math for anyone who previously hit a wall mid-conversation and was asked to upgrade. OpenAI is absorbing the inference cost of uncapped text for a far larger audience, a bet that broader access keeps users inside its ecosystem.
The same update includes safety work. OpenAI tightened the model's behavior for users under 18, with new guardrails that block romantic roleplay and enforce age-appropriate boundaries, layered on the protections already in place.
Together, Luna and Sol show how OpenAI is dividing the GPT-5.6 generation: speed for the free tier, frontier capability for paying customers. The changes arrive in two stages, the model switch this week and unlimited text next week, so free users get Luna as their default before the caps come off.
Why this matters
With the ChatGPT free text chat limits gone, free users get a current-generation default model and no cap on plain text conversation, which makes the free tier materially more useful for daily use. The most visible reason to upgrade for people who only wanted more chat volume is gone. The factual-error numbers and the under-18 safety work are the details to watch as the rollout completes over the next two weeks.
Sources
Improving GPT-5.6 Sol in ChatGPT—and expanding access to GPT-5.6 Luna for free users
Improving GPT‑5.6 Sol in ChatGPT—and expanding ...
GPT-5.6: Frontier intelligence that scales with your ambition
Photo by Brecht Corbeel on Unsplash
Related Articles
- GPT-5.6 Government Review: OpenAI Launches Sol, Terra, Luna
- GPT-5.6 on Amazon Bedrock: OpenAI's Sol, Terra, and Luna Models Arrive for Enterprise Deployments
- OpenAI Launches GPT-5.5 with Drastic Cost Cuts and Advanced Agentic Capabilities
✔Human Verified
Researched and cross-referenced against primary sources by the Bytevyte editorial team. This article was generated with the assistance of artificial intelligence and reviewed by the Bytevyte editorial team.