OpenAI Removes Text Message Rate Limits for Free ChatGPT Accounts

OpenAI will stop enforcing rate limits on text-only prompts for ChatGPT Free and Go tier users starting next week, the company announced. The change means users on those plans can send unlimited text messages to ChatGPT without hitting a usage cap — a first for OpenAI's free offering.

The removal applies specifically to text-based conversations. OpenAI will continue to enforce separate limits for prompts that include files or images, as well as for image generation and ChatGPT's recently updated voice mode. Combining text chats with these other features will count against those separate limits.

Alongside the rate limit change, OpenAI is making GPT-5.6 Luna — the smallest model in its GPT-5.6 family — the new default model for Free and Go accounts. The switch takes effect this week and replaces GPT-5.5 Instant, which had served as the default since May 2026. A new "Think" button will let Free and Go users prompt GPT-5.6 Luna to spend additional time generating responses. That feature is also unlimited for pure text chats, but usage counts toward a separate limit if combined with other inputs.

Plus and Pro subscribers are getting their own updates. OpenAI is refreshing GPT-5.6 Sol, its current flagship model, to better handle everyday conversation. According to OpenAI, the updated system "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." Plus and Pro users will also get a new slider — similar to Anthropic's effort control in Claude — that lets them adjust how much computational "thought" GPT-5.6 Sol applies to a given answer.

OpenAI did not disclose the technical mechanism behind the rate limit removal. Rate limits typically reflect inference costs — what AI providers pay to run trained models against incoming requests — so eliminating them for free-tier text chats implies a meaningful reduction in per-query cost, cost absorption, or infrastructure capacity increase. A recent report from The Information reported that OpenAI was nearing an inference-related breakthrough, though OpenAI has not confirmed a direct connection between that report and this rollout.

In a statement, OpenAI framed the move as part of a broader push toward what it called "abundant intelligence": "This is a concrete step toward more abundant intelligence: making our latest models more widely available, improving the usefulness and reliability of the answers people get, and letting free users keep text chats going without a rate limit," the company said. "Access shapes opportunity, and this update gives more people the ability to keep asking, develop an idea, and get help when they need it."

What this means

Removing rate limits on text chats — even with restrictions on multimodal features intact — signals that OpenAI's per-query inference costs for smaller models like GPT-5.6 Luna have dropped enough to sustain unlimited free usage at scale. This is a meaningful competitive move against rivals like Google's Gemini and Anthropic's Claude, both of which maintain usage caps on free tiers. It also raises questions about monetization: if unlimited text access no longer differentiates paid tiers, OpenAI is betting that multimodal features, model quality (via GPT-5.6 Sol), and the new effort slider will be enough to justify Plus and Pro subscriptions. Whether this pattern extends to other model sizes, or whether OpenAI can sustain it if usage surges, remains to be seen.