TechCrunch reported Thursday that OpenAI will remove text-chat rate limits for ChatGPT Free and Go users and make GPT-5.6 Luna their default model. The text-only change pairs the smaller model with a Think button that gives it more time on harder prompts. The expansion puts a current-generation model in front of a service the company says crossed one billion weekly users.

What Changed

AI-generated summary, reviewed by an editor. More on our AI guidelines.

Free access

OpenAI's announcement puts Luna in place of GPT-5.5 Instant during the week of August 6, with unlimited text chats and the Think button due the following week. Caps stay on file uploads, images, voice conversations and other tools, and the changes apply to the ChatGPT apps for iPhone, iPad and Mac.

Luna is the speed-focused lightweight model in the three-model GPT-5.6 lineup. The Think button does not move a request to the flagship Sol model; it keeps the conversation on Luna and gives it more processing time.

Plus and Pro subscribers gained access to an updated version of GPT-5.6 Sol and a reasoning slider on the day of the August 6 announcement. The slider, reached through a speedometer icon, offers settings from Instant through Pro. The ChatGPT update does not change the Sol versions used in Codex or ChatGPT Work.

Go costs $8 a month, Plus $20 and Pro $200. Unlimited text access on the free tiers is still subject to abuse guardrails.

OpenAI described the revised Sol as more direct and better able to adjust the detail of an answer to the question, and said the model is designed to make fewer mistakes when answers depend on dates, numbers, sources, rules or assumptions. The slider lets paid users decide “how much thought ChatGPT puts into an answer.”

Accuracy claims

OpenAI's internal evaluation found that responses containing at least one factual error were about 62 percent less common with Luna and 68 percent less common with the updated Sol than with GPT-5.5 Instant. The tests covered prompts involving finance, medicine and law. Those figures come from OpenAI's own evaluation and had not been independently verified as of August 7.

Know someone who'd find this useful? ✉️ Email it to a friend in one click, or they can subscribe free here.

The tier dispute

The Decoder published its report under the headline “OpenAI improves GPT-5.6 Sol in ChatGPT and restricts free users to its weakest model.” OpenAI contacted the outlet after publication and said it was “not accurate to represent this as free users being restricted to a weaker experience.” The company added that free users' capabilities “have only grown compared to what was available before.”

The outlet amended its story and added an August 7 update. Matthias Bastian wrote that his core point still stood: “The free tier has improved in absolute terms, but free users no longer get any access to OpenAI's frontier reasoning models.” He noted that the Think button lets Luna work longer but does not switch the user to a stronger model.

Bastian also questioned whether more controls would help, pointing to users' limited interest in an earlier model switcher. The difference among tiers is easy to miss, he wrote: “Many people don't realize this because they assume 'ChatGPT is ChatGPT.'”

Frequently Asked Questions

What exactly becomes unlimited for free ChatGPT users?

Text chats only. Free and Go users will no longer hit rate limits on text-based conversations, but caps remain on file uploads, images, voice conversations and other tools. The unlimited text access is still subject to abuse guardrails.

Which model do free users get, and when?

GPT-5.6 Luna replaces GPT-5.5 Instant as the default for Free and Go users during the week of the August 6 announcement. Unlimited text chats and the Think button are due the following week.

Does the Think button give free users access to the flagship model?

No. According to The Decoder, the Think button keeps the request on Luna and allows more processing time. It does not move a user to the flagship GPT-5.6 Sol model.

What do Plus and Pro subscribers get?

An updated version of GPT-5.6 Sol and a reasoning slider, both available on the day of the August 6 announcement. The slider is reached through a speedometer icon and offers settings from Instant through Pro. The Sol versions used in Codex and ChatGPT Work are unchanged.

How reliable are the accuracy figures OpenAI cited?

They come from OpenAI's own internal evaluation, which covered prompts involving finance, medicine and law. The company reported that responses containing at least one factual error were about 62 percent less common with Luna and 68 percent less common with the updated Sol than with GPT-5.5 Instant. The figures had not been independently verified as of August 7.

AI-generated summary, reviewed by an editor. More on our AI guidelines.

Commerce Department Clears OpenAI's GPT-5.6 for Broad Public Release
The U.S. Department of Commerce has cleared OpenAI's GPT-5.6 model family for broad public release, a person familiar with the matter told Axios on Tuesday. OpenAI announced late Tuesday night that th
The Implicator
OpenAI Releases GPT-5.6 Broadly and Launches ChatGPT Work Agent
OpenAI released its GPT-5.6 model family to the public on Thursday and introduced ChatGPT Work, an agent that gathers context from a user's apps and files to produce documents, spreadsheets, presentat
The Implicator
Meta Prices Its First Paid AI Model API at $4.25 per Million Output Tokens
Meta Superintelligence Labs released Muse Spark 1.1 on Thursday and opened a public preview of the Meta Model API, the first time the company has charged outside developers for access to a model it bu
The Implicator
AI News

San Francisco

Editor-in-Chief and founder of Implicator.ai. Former ARD correspondent and senior broadcast journalist with 10+ years covering tech. Writes daily briefings on policy and market developments. Based in San Francisco. E-mail: editor@implicator.ai