ChatGPT Removes Free-Tier Chat Limits: What Changed, and Why

OpenAI announced on August 6, 2026 that free ChatGPT users are getting unlimited text chats, removing one of the last hard caps on its no-cost tier. The change runs on a newly-elevated GPT-5.6 Luna as the default free model, paired with a "Think" button that lets free users opt into deeper reasoning on demand. Paying Plus and Pro subscribers get their own upgrade the same week: a reasoning-effort slider for GPT-5.6 Sol.

Free-tier limits are the thing most people actually notice about ChatGPT day to day, so this is a bigger deal than a typical model bump. If you've ever hit a "you've reached your limit, try again later" wall mid-conversation, that specific friction is going away. Here's exactly what's changing, why OpenAI is doing it now, and how the no-cost tier stacks up against Claude and Gemini once the caps come off.

Quick Take

• What changed: Free and Go-tier ChatGPT users get unlimited text conversations, powered by GPT-5.6 Luna as the new default model, plus a "Think" button to trigger deeper reasoning on individual questions.
• What didn't change: Limits on file uploads, image generation, and voice mode stay in place for free users — this is specifically about text chat volume.
• For paying users: Plus and Pro subscribers get a reasoning-effort slider on GPT-5.6 Sol, letting them dial reasoning up or down per question instead of getting one fixed setting.
• Rollout: The Plus/Pro slider and upgraded Sol started rolling out Thursday, August 6. Unlimited free chats and the Think button are following the week after.
• Why it matters: It's a direct answer to Claude's and Gemini's own free tiers, which both still cap daily usage — OpenAI is betting that removing friction on the free tier keeps more people from ever trying a competitor.

What Actually Shipped

The core change is simple to state: free and Go-tier ChatGPT users no longer have a cap on how much they can type into a text conversation. Until now, free accounts ran into periodic "you've hit your limit" messages that pushed people either to wait it out or upgrade. That wall is coming down, with GPT-5.6 Luna — OpenAI's fastest, cheapest model in the GPT-5.6 family — running as the default engine behind it.

Alongside unlimited chat, free users get a new Think button, letting them opt a single hard question into deeper reasoning rather than running every reply at the same fixed depth — a lighter version of the reasoning controls previously reserved for paying tiers.

Plus and Pro subscribers get their own upgrade: an improved GPT-5.6 Sol plus a reasoning-effort slider across web, mobile, and desktop, so you can dial reasoning down for a quick fact-check or up for research, planning, or a gnarly coding problem. OpenAI also reports lower factual error rates for both Luna and Sol versus the prior GPT-5.5-Instant generation — its own internal measurement, not an independently audited benchmark, so treat it as directional rather than a hard number.

One thing worth being precise about: "unlimited" applies to text chat specifically. Free accounts still face separate, unchanged caps on file uploads, image generation, and voice mode — those are the parts of ChatGPT that cost OpenAI meaningfully more to run per request, and they're staying metered.

Why OpenAI Is Doing This Now

Removing limits on your free tier is an unusual move for a company also reportedly burning through massive compute costs, and it only makes sense in the context of the last few weeks. OpenAI cut API prices on GPT-5.6 Luna by 80% just one day earlier, on August 5 — a move it attributed to efficiency gains in the model and its serving infrastructure. Running Luna cheaper at the infrastructure level is very likely what made an "unlimited" free tier financially viable in the first place; the two announcements landing back to back is not a coincidence.

There's also a clear competitive angle. Both Claude's and Gemini's free tiers still ration usage: Claude's free plan is widely reported to run on a rolling window of roughly 15–40 messages per five hours, and Gemini's free tier is generally understood to cap out around 30 prompts a day, though Google doesn't publish an exact number. Genuinely unlimited text chat is something neither competitor currently offers on their free plan. For a company OpenAI has said has crossed a billion active users, keeping the free experience frictionless is a straightforward way to keep people from ever having a reason to open a rival app.

How This Changes the Free-Tier Picture

For casual users — people asking ChatGPT for advice, quick research, or help drafting something once or twice a day — this removes the one thing that used to force a decision between paying $20/month for Plus or living with interruptions. That's a meaningful shift in how the three major assistants compete at the entry level, a dynamic we track in more depth in our Claude vs. ChatGPT vs. Gemini comparison.

It's worth being clear-eyed about what "unlimited" doesn't fix, though. Free users still get GPT-5.6 Luna by default rather than the more capable Sol, so anyone doing serious coding, long-document analysis, or research work will hit a real quality ceiling well before a message cap — the Think button narrows that gap for individual hard questions, but it doesn't turn Luna into Sol. If your use case leans toward demanding work, our breakdown of AI tools actually worth paying for in 2026 is a better guide than chasing the "free and unlimited" headline. Teams comparing coworker-style AI subscriptions rather than plain chat face a different set of trade-offs — see how ChatGPT Work stacks up against Claude Cowork and Gemini Spark.

The slider on the paid side is arguably the more useful change for power users: fixed reasoning depth has been a real annoyance since OpenAI first split GPT-5.6 into Sol, Terra, and Luna tiers, since paying for Sol only helped if OpenAI's own default reasoning setting matched what a given question actually needed. Being able to turn it down for a quick question, or up for something that requires the model to check its own work, makes the $20/month Plus tier meaningfully more useful without a price change — a contrast worth keeping in mind next to Claude Opus 5's own recent pricing moves.

Frequently Asked Questions

Does "unlimited" mean everything in ChatGPT is free and uncapped now? No. Only text chat is uncapped for free and Go-tier users. File uploads, image generation, and voice mode all keep their existing, separate limits.

Which model do free users get now? GPT-5.6 Luna, OpenAI's fastest and cheapest model in the GPT-5.6 lineup, becomes the default for Free and Go tiers. It's a capable everyday model but not the same tier as Sol, which stays reserved for Plus, Pro, Business, and Enterprise.

What is the Think button? A toggle free users can hit on a specific question to have ChatGPT apply deeper reasoning to that one answer, instead of running every reply at the same fixed depth.

What do Plus and Pro subscribers actually get out of this? An upgraded GPT-5.6 Sol plus a reasoning-effort slider they can adjust per question — lower for quick answers, higher for research, planning, or complex coding tasks.

When does this roll out? The Sol update and reasoning slider for Plus/Pro began rolling out Thursday, August 6, 2026. Unlimited free chats and the Think button for Free/Go users are rolling out the following week.

Is ChatGPT's free tier now more generous than Claude's or Gemini's? On raw text chat volume, yes — neither Claude's nor Gemini's free plans currently offer unlimited daily messaging. Both still cap heavier features like advanced reasoning, research tools, and image generation on their free tiers too, so the comparison isn't perfectly apples-to-apples, but ChatGPT's move closes a real gap that used to favor upgrading.

Bottom Line

Removing the free-tier chat cap is a genuinely user-favorable change, and it's a direct shot at Claude's and Gemini's still-limited free plans — but "unlimited" here means unlimited text chat on a lower-tier model, not unlimited access to everything ChatGPT can do. Casual users get real, immediate relief from an annoying limit; anyone doing demanding work will still find good reasons to pay. Watch whether Anthropic and Google respond with looser free tiers of their own in the coming weeks — that's usually how these moves play out. For the fuller picture of how the three major assistants compare beyond pricing and limits, read our Claude vs. ChatGPT vs. Gemini breakdown.