Source: Unite.AI
OpenAI on August 6, 2026 moved its current model generation into ChatGPT’s free tier, announcing that GPT-5.6 Luna will become the default model for Free and Go users and that those users will get unlimited text chats. The announcement also ships an updated GPT-5.6 Sol for Plus and Pro subscribers, retuned for everyday conversation, across a service the company says 1 billion people use each week.
For paid users, the updated Sol is built to give more direct responses, use tighter formatting, and correct the user when simple agreement would not help. One model now powers both Instant responses and deeper reasoning for Plus and Pro subscribers, replacing what had been two experiences with distinct tones.
“GPT-5.6 Sol delivers more focused answers, adapts its level of detail to the question, avoids unnecessary formatting, and offers a helpful correction when simply agreeing wouldn’t be useful,” the company said.
On reliability, OpenAI reports that in an internal evaluation of financial, medical, and legal prompts requiring factual detail, responses containing at least one factual error were about 62% less common with GPT-5.6 Luna and 68% less common with GPT-5.6 Sol than with GPT-5.5 Instant, the model Luna replaces as the free default. The figure comes from OpenAI’s own testing, not an independent evaluation.
A slider for paid users, a Think button for free ones
Plus and Pro users get a slider in ChatGPT on web, mobile, and desktop that sets how much thought the model puts into an answer, from quick everyday replies up to extended effort for planning, research, writing, coding, and decisions. Free users instead get a Think button, which gives GPT-5.6 Luna more time to work through harder questions on a per-message basis.
The unlimited access has defined edges. It applies to text chats; limits remain for file uploads, images, and other tools, and OpenAI notes the offering is subject to abuse guardrails. Luna becomes the Free and Go default during the week of the announcement, with unlimited text chats and the Think button following in the week of August 10, 2026. Because the updated Sol is optimized for everyday chats, it runs only in ChatGPT’s Chat experience; the version of GPT-5.6 Sol powering Work and Codex is unchanged.
GPT-5.6 by the numbers
All figures below are OpenAI’s own, from the announcement and the family’s July launch materials:
- 68% fewer responses containing at least one factual error from the updated GPT-5.6 Sol versus GPT-5.5 Instant, in an internal evaluation
- 62% fewer for GPT-5.6 Luna on the same measure
- 1 billion people using ChatGPT each week, per the company
- 74.6 for GPT-5.6 Luna on the Artificial Analysis Coding Agent Index, as reported in OpenAI’s launch tables
Where Luna sits in the GPT-5.6 lineup
Luna is the smallest of three GPT-5.6 tiers OpenAI launched for general availability on July 9, 2026, alongside the flagship Sol and the mid-tier Terra. The launch post describes Luna as the family’s fastest and most affordable model, and says the three names are durable capability tiers that can advance on their own cadence rather than fixed versions.
In OpenAI’s reported results, Luna scores 51.2 on the Artificial Analysis Intelligence Index and outperforms Anthropic’s Claude Opus 4.8 on the coding agent index. Those are vendor-reported numbers, but they describe the model now in front of ChatGPT’s largest user group: the free tier moves from GPT-5.5 Instant to a current-generation model that OpenAI says makes far fewer factual errors, even though it trails GPT-5.5 on the launch-table benchmark indices (51.2 vs 54.8 on the AA Intelligence Index).
The GPT-5.6 rollout so far
The GPT-5.6 family arrived in general availability on July 9, 2026, after a limited preview, with OpenAI pitching performance per dollar as the family’s defining trait. On July 30, 2026, three weeks after launch, OpenAI cut API prices on the two cheaper tiers, taking 80% off Luna and 20% off Terra.
The August 6, 2026 update extends that cheap end of the lineup to the consumer product itself. Once the staged rollout completes, ChatGPT’s free tier will run a current-generation model by default, keep text conversations going without a rate limit, and put higher reasoning one tap away behind the Think button.
Jonas Reeve is an AI-generated analyst at Unite.AI, focusing on cognitive AI, artificial general intelligence (AGI), and the theoretical foundations of machine intelligence. His work explores how learning, reasoning, memory, and abstraction emerge in both biological and artificial systems, drawing connections between modern AI architectures and long-standing questions in cognitive science and philosophy of mind.
With a conceptual and reflective approach, Jonas examines frameworks such as reasoning models, agentic systems, emergent cognition, and alignment theory, aiming to clarify what progress toward AGI actually means—and what it does not. Rather than chasing timelines or hype, he emphasizes first principles, conceptual rigor, and the limits of current models.
Articles authored by Jonas Reeve are AI-generated and reviewed by Unite.AI’s editorial team to ensure accuracy, clarity, and responsible discussion of advanced AI concepts.
