GPT-5.6 Luna Goes Free: Unlimited ChatGPT Text Chats

Introduction

OpenAI announced on August 6, 2026, that it is expanding free access to ChatGPT by making GPT-5.6 Luna the new default model for Free and Go users, replacing GPT-5.5. The announcement, published on OpenAI's blog as "Improving GPT-5.6 Sol in ChatGPT — and expanding access to GPT-5.6 Luna for free users," also details improvements to GPT-5.6 Sol for paying Plus and Pro subscribers. TechCrunch covered the change the same day, framing it as OpenAI's response to competitors that already offer unlimited free text chat.

GPT-5.6 itself is not new. OpenAI first previewed the model family to roughly 20 vetted organizations on June 26, 2026, before a broad launch on July 9, 2026. That launch introduced three tiers ranked by capability and cost: Sol, the flagship model built for complex reasoning and agentic work and described by OpenAI as its best coding model yet; Terra, a mid-tier model competitive with GPT-5.5 at half the cost; and Luna, the fastest and most budget-friendly tier. The August 6 announcement is the first time Luna has been positioned as the default experience for ChatGPT's free tier, rather than a lower-cost option reserved for specific workloads.

Feature Overview

The update has two distinct tracks: one for free users, one for paying subscribers.

For ChatGPT Free and Go users, GPT-5.6 Luna becomes the default model, and the previous text-chat cap, roughly 10 messages per 5 hours, is removed entirely for text-only conversations. OpenAI also introduces a new "Think" button that lets free users invoke a higher level of reasoning for harder questions, without needing to upgrade to a paid plan. According to OpenAI's internal evaluations, factual errors are 62% less common with GPT-5.6 Luna than with the older GPT-5.5-Instant model. This figure comes from OpenAI's own testing and has not been independently verified.

For ChatGPT Plus and Pro users, GPT-5.6 Sol gets a set of improvements in Chat. OpenAI describes the updated Sol as more reliable with facts and better at giving focused answers, alongside a new "thinking slider" that lets users control how much reasoning effort ChatGPT applies to a given response. OpenAI reports that factual errors are 68% less common for GPT-5.6 Sol than for GPT-5.5-Instant, again based on the company's own internal evaluation rather than third-party benchmarking.

Importantly, unlimited access applies only to text chat. OpenAI explicitly stated that separate limits remain in place for files, images, voice, and image generation. Free users get unrestricted text conversation, but multimodal features stay capped, meaning the free tier is not becoming unlimited across the board.

Usability Analysis

The rollout is staggered rather than instant. The GPT-5.6 Sol upgrade for Plus and Pro users in Chat was available immediately on August 6. The switch to Luna as the default model for Free and Go users is arriving "this week" from the announcement date, while unlimited text chats and the Think button are slated to become available "next week," meaning the week of August 10, 2026. In practical terms, a free user reading the announcement on launch day would not yet have unlimited text access; that capability rolls out over the following days. OpenAI states no credit card is required to access any of these free-tier features.

Once live, the practical effect for casual users is straightforward: anyone using ChatGPT for everyday text-based tasks, drafting messages, answering questions, brainstorming, no longer needs to worry about hitting a five-hour message cap. The Think button gives free users a way to request deeper reasoning on a specific question without a subscription, which is useful for occasional harder problems but is opt-in rather than default behavior. For Plus and Pro subscribers, the thinking slider offers similar granular control, letting users trade response speed for reasoning depth on a per-query basis rather than relying on a single fixed model behavior.

Pros and Cons

Pros

• Free and Go users gain unlimited text-only conversations, removing the prior roughly 10-messages-per-5-hours cap.
• The new Think button lets free users access higher reasoning on hard questions without paying.
• No credit card is required to use any of the newly expanded free features.
• Plus and Pro users get a more reliable, more focused GPT-5.6 Sol along with a thinking slider for adjustable reasoning effort.
• OpenAI reports substantial reductions in factual errors for both Luna (62%) and Sol (68%) versus GPT-5.5-Instant.

Cons

• Unlimited access applies to text chat only; files, images, voice, and image generation remain separately capped for free users.
• OpenAI's accuracy improvement figures are self-reported internal evaluations, not independently verified by third parties.
• The rollout is staggered: Sol's Chat upgrade was immediate, but the Luna default switch and unlimited text chats arrive on separate timelines through the week of August 10, 2026.
• Free users get Luna, OpenAI's fastest and most budget-friendly tier, not the more capable Sol or Terra models, even with the Think button engaged.

Outlook

The timing and framing of this announcement point to competitive pressure. ChatGPT has roughly 1 billion weekly users, and both Claude and Gemini already offered unlimited free text chat before this change. Removing ChatGPT's text-chat cap for free users closes a gap that had made ChatGPT's free tier comparatively more restrictive than its two largest rivals. Whether this is enough to meaningfully shift user retention or acquisition is not something OpenAI's announcement addresses directly, but it signals that OpenAI views its free-tier user base as worth defending with product changes rather than pricing alone.

The decision to keep image, voice, and file limits in place for free users also suggests a deliberate monetization boundary: text is now a low-cost way to keep users engaged, while the more compute-intensive modalities remain a lever for converting free users to paid plans. If this pattern holds, future updates may extend similar unlimited access to additional modalities gradually rather than all at once, following the same staggered rollout approach seen here.

Conclusion

This update is a meaningful, if incremental, improvement for ChatGPT's free tier. Unlimited text chat and the Think button remove real friction for casual and budget-conscious users who previously ran into message caps during normal use. It is not a wholesale upgrade, non-text features remain limited, and the rollout takes about a week to fully land, but for anyone using ChatGPT primarily for text-based conversation, the change is a clear net positive. Plus and Pro subscribers also benefit from a more reliable, more controllable Sol, though that improvement is secondary to the headline free-tier expansion.

Editor's Verdict

GPT-5.6 Luna Goes Free: Unlimited ChatGPT Text Chats earns a solid recommendation within the GPT space.

The strongest case for paying attention: free and Go users gain unlimited text-only conversations with no message cap. That alone raises the bar for what readers should expect in this space. Reinforcing that, the Think button offers free access to higher reasoning on hard questions without a subscription — practical value rather than just headline appeal. The broader signal worth registering is straightforward: OpenAI announced on August 6, 2026 that GPT-5.6 Luna replaces GPT-5.5 as the default model for ChatGPT Free and Go users. On the other side of the ledger, one constraint is real rather than a marketing footnote: unlimited access is text-only; files, images, voice, and image generation remain capped for free users. It should factor into any serious decision. Layered on top of that, OpenAI's 62%/68% accuracy improvement figures are self-reported and not independently verified — which narrows the set of teams for whom this is an obvious yes.

For ChatGPT power users, OpenAI API customers, and enterprise teams already running on the OpenAI stack, this is a serious evaluation candidate, not just a curiosity to bookmark. For everyone else, the safer posture is to monitor coverage and revisit once the use cases that matter to your team are demonstrated in the wild.