OpenAI makes GPT-5.6 Luna the default for free ChatGPT users
Free and Go users will get unlimited everyday chats and a Think button, while Plus and Pro users get an updated Sol model with adjustable reasoning.
By Ryan Merket · Published · Updated
Why it matters
OpenAI is using Luna's lower inference cost to remove a major free-tier restriction. Paid ChatGPT is increasingly differentiated by model quality, reasoning control and tools rather than basic chat access.
OpenAI will make GPT-5.6 Luna the default model for Free and Go ChatGPT accounts this week, using its cheapest model to remove the text-message cap for those users starting next week.
The August 6th announcement also splits ChatGPT's consumer experience more clearly by subscription tier. Free and Go users will receive unlimited Luna text chats and a new Think button for harder questions. Plus and Pro subscribers get an updated version of GPT-5.6 Sol, along with a slider that controls how much reasoning time ChatGPT applies to a response.
OpenAI is turning a rapid drop in inference costs into wider distribution. GPT-5.6 became generally available on July 9th, and OpenAI cut Luna's API price by 80% on July 30th. One week later, Luna is becoming the engine behind ChatGPT's broadest and most price-sensitive user segment.
Axios first reported the consumer changes, which OpenAI announced on Thursday.
What changes for free ChatGPT users
Luna will replace GPT-5.5 Instant as the default model for Free and Go accounts this week. Unlimited text chats and the Think button are scheduled to follow next week, subject to abuse controls.
The unlimited policy applies only to text. OpenAI said limits will remain for file uploads, image features and other tools, preserving some of the usage boundaries that distinguish paid subscriptions.
Think gives Luna additional time to reason through a prompt. That puts a visible reasoning control into the free product without giving those accounts access to the higher-end Sol model used by Plus and Pro subscribers.
OpenAI says an internal evaluation covering financial, medical and legal questions found that answers containing at least one factual error were about 62% less common with Luna than with GPT-5.5 Instant. OpenAI did not publish enough detail in the announcement to independently assess the test's prompt set or scoring process, so the result remains a company-reported benchmark.
OpenAI reshapes the paid tier around Sol
Plus and Pro users can access the revised GPT-5.6 Sol starting August 6th on ChatGPT's web, mobile and desktop products. OpenAI says the model now gives more focused answers, uses less unnecessary formatting and adjusts the level of detail to the question.
The update also brings Instant and deeper-reasoning responses under the same Sol model. The new slider lets users keep reasoning low for quick questions or increase it for planning, research, writing, coding and other complex work.
OpenAI reported a 68% reduction in responses containing at least one factual error for the updated Sol model compared with GPT-5.5 Instant, using the same internal evaluation cited for Luna.
The change applies specifically to the Chat experience. OpenAI said the GPT-5.6 Sol versions running inside ChatGPT Work and Codex are unchanged, separating the conversational tuning announced Thursday from the models used for longer agentic and software-development tasks.
Cheaper inference becomes a distribution tool
OpenAI launched the GPT-5.6 family with three tiers: Sol for its highest-capability work, Terra for a balance of performance and cost, and Luna for fast, high-volume workloads. OpenAI initially priced Luna at $1 per million input tokens and $6 per million output tokens.
On July 30th, OpenAI reduced those prices to $0.20 per million input tokens and $1.20 per million output tokens. Terra received a smaller 20% reduction, while Sol pricing remained unchanged.
That sequence matters. OpenAI first established Luna as the low-cost member of its newest model family, then cut its price weeks after launch, and is now using it to eliminate a central restriction on free ChatGPT usage. The lower serving cost gives OpenAI room to increase engagement without putting its most expensive model behind every free conversation.
Paid differentiation is moving toward model quality, reasoning control and access to tool-heavy products. Free users receive a capable default model with effectively open-ended text use. Plus and Pro subscribers receive Sol, an adjustable reasoning budget and the wider set of paid features attached to their plans.
OpenAI says 1 billion people use ChatGPT each week. At that scale, making Luna the free default is also a test of whether efficiency gains can translate into materially broader usage while keeping the higher-cost Sol experience valuable enough to support subscriptions.