OpenAI removes rate limits on ChatGPT text chats across every tier

ChatGPT text chats are now unlimited, and the free tier is the one that changed most.

What changed on August 6

OpenAI removed the rate limits on text-based conversations for every tier on August 6, 2026. Free and Go users also move to GPT-5.6 Luna as the new default model, replacing GPT-5.5 Instant, and get a Think button that lets the model spend longer on a hard question instead of answering fast. Plus and Pro users get an upgraded GPT-5.6 Sol plus a slider controlling how much thinking they want.

The accuracy claims come from OpenAI’s own internal evaluations: factual errors were 62% less common with GPT-5.6 Luna and 68% less common with GPT-5.6 Sol, both measured against GPT-5.5 Instant. Internal numbers deserve the usual caveat, a direction rather than a guarantee.

What did not change

Unlimited means text. Files, images, voice and image generation all keep their separate limits.

The rollout is staged. Sol landed for Plus and Pro on the 6th, the Free and Go changes roll out through the week, and unlimited chats plus the Think button arrive the following week.

The infrastructure statement

ChatGPT crossed 1 billion weekly users shortly before the announcement, and that number is the real reason this move is interesting. Removing rate limits at that scale is an infrastructure statement, not a feature. Rate limits exist to protect capacity; dropping them for the entire text workload, including the free tier, signals that OpenAI believes its serving capacity can absorb whatever a billion weekly users send at it.

The free tier arguably gains the most: a newer default model, a reasoning button that was previously a paid-tier signature, and no ceiling on conversations. For the paid tiers, the upgrade is finer control rather than more quota, which suggests where OpenAI now sees the differentiation between free and paid living: in how much thinking you can dial up, not in how often you can ask.