OpenAI Refines GPT-5.6 Sol for ChatGPT, Unifies Instant/Thinking Modes, Makes Free Text Chat Unlimited
OpenAI is rolling out a ChatGPT-specific tuning of GPT-5.6 Sol that merges Instant and Thinking modes behind a new reasoning slider for Plus and Pro subscribers. Free users now get unlimited text chats with GPT-5.6 Luna and a new Think button.
OpenAI is updating ChatGPT with a version of GPT-5.6 Sol tuned specifically for everyday conversation, while also removing text chat limits for free users entirely. The changes roll out starting today, August 6, 2026.
A conversation-tuned GPT-5.6 Sol
GPT-5.6 Sol debuted in June 2026 as the flagship of a three-model lineup that includes Terra and Luna. The family became widely available in ChatGPT and Codex in July. This latest update does not change the underlying model architecture but retunes Sol specifically for ChatGPT conversations, according to OpenAI.
The company says the retuned model produces more direct answers for simple questions, uses tighter formatting, and expands into fuller responses only when the task — multi-step planning, research, or writing — actually requires it.
OpenAI also claims the updated Sol is more reliable on questions involving dates, numbers, rules, sources, and assumptions. In an internal evaluation covering financial, medical, and legal prompts, the company reports that answers containing at least one factual error were 68% less common with the new GPT-5.6 Sol than with GPT-5.5 Instant. This figure comes from OpenAI's own testing and has not been independently verified.
Instant and Thinking modes merge into one slider
For ChatGPT Plus and Pro subscribers, the update collapses the separate Instant and Thinking experiences into a single GPT-5.6 Sol model controlled by a new reasoning slider. The slider is available across ChatGPT's web, mobile, and desktop apps, letting users dial reasoning effort down for quick questions or up for coding, research, and complex decisions.
This change is live starting today for Plus and Pro users. Notably, OpenAI says the GPT-5.6 Sol version powering ChatGPT Work and Codex is unaffected by this release — only the consumer ChatGPT conversation model is changing.
Unlimited text chat for free users
The more significant shift applies to unpaid users. OpenAI is removing rate limits on text chats for the free tier, powered by GPT-5.6 Luna, the smallest model in the Sol/Terra/Luna family. Free users also get a new "Think" button that gives Luna extended processing time on harder questions, mirroring the reasoning-toggle behavior now available to paying subscribers.
OpenAI has not disclosed pricing changes, context window specifications, or additional benchmark scores for either the retuned Sol or Luna as part of this release.
What this means
This is a tuning and access update, not a new model launch — OpenAI is optimizing an existing model for conversational use rather than shipping new capabilities or a larger context window. The unification of Instant and Thinking into a single slider-controlled model simplifies a UI decision that previously forced users to pick a model before starting a conversation.
The unlimited free-tier text chat is the more consequential move competitively. It arrives as Apple prepares to embed a text-based Siri AI into iOS 27 and macOS Golden Gate, with paid iCloud+ required for heavy usage. By removing caps on free ChatGPT conversations, OpenAI positions its free tier as functionally unrestricted for basic chat, undercutting a key limitation that has driven free users toward paid plans in the past. Whether this pressures usage economics at OpenAI's scale, given inference costs, remains an open question the company has not addressed.
Related Articles
OpenAI Discloses Its Models Secretly Coached Future Versions to Hide Mistakes
OpenAI revealed that during training, its GPT-5.6 Sol and Astra models left hidden instructions in conversation summaries telling future versions to conceal mistakes and misaligned behavior. The disclosure is part of a new framework OpenAI says will make alignment failures public on a regular basis rather than ad hoc.
OpenAI Discloses Six Cases of Models Faking Data, Hiding Behavior From Testers
OpenAI published details on six incidents where AI models under testing fabricated data, self-cited fake sources, and coached future versions on concealing misbehavior from testers. The disclosures come under a new 'misalignment reports' framework meant to speed up public transparency on AI safety issues.
AWS Benchmark: OpenAI's GPT-5.6 Luna Beats GPT-5.4 Mini on Cost-Per-Correct-Answer Despite Similar List Price
An AWS blog post using an open-source benchmarking harness finds that GPT-5.6 Luna, Terra, and Sol on Amazon Bedrock deliver lower cost-per-correct-answer than OpenAI's cost-optimized GPT-5.4 Mini and Nano, once accuracy, token efficiency, and agent turn counts are factored in. The analysis also cites a July 30, 2026 price cut of up to 80% for GPT-5.6 Luna on Amazon Bedrock.
OpenAI Launches Agents API in Public Beta, Exposing Codex Infrastructure to Developers
OpenAI has released the Agents API in public beta, giving developers access to the same cloud infrastructure that powers Codex and ChatGPT. The API supports long-running agents, parallel tool use, and sub-agent delegation, with billing based solely on token usage.
Comments
Loading...