OpenAI is removing text-chat limits for all ChatGPT users, free tier included. The announcement came today, not long after ChatGPT crossed 1 billion weekly users.

What's changing

Free and Go users get a new default model, GPT-5.6 Luna, replacing GPT-5.5. Both tiers also get a "Think" button for requesting higher reasoning power on complex questions. Limits still apply to files, images, voice, and image generation — this change is specifically about text chat.

Plus and Pro subscribers get an upgraded GPT-5.6 Sol, tuned for quick tasks — questions, web research, advice, planning, writing, decision-making — with more compact, direct answers. It's a separate model version from the GPT-5.6 Sol used in Codex and Work, which stays unchanged.

Plus and Pro users also get a thinking slider, so they can dial reasoning effort up or down depending on how complex the query actually is.

The numbers OpenAI is citing

OpenAI says an internal evaluation found factual errors dropped 62% for GPT-5.6 Luna and 68% for GPT-5.6 Sol, measured against GPT-5.5-Instant. Worth flagging: that's OpenAI grading its own model. Not a reason to dismiss it, just a reason not to treat it as independently verified.

Rollout timing

The updated GPT-5.6 Sol is live now for Plus and Pro. The Free and Go changes — new Luna model, Think button — start rolling out this week. Unlimited text chats for Free and Go land next week.

Why this matters

Dropping usage caps on the free tier is a real move in how OpenAI positions ChatGPT against competitors already offering generous free access. And pairing that with a reasoning toggle on paid tiers, instead of one fixed model, points at where this is headed: reasoning effort becomes something you dial per query, not something baked into which tier you're paying for.

Source: TechCrunch — ChatGPT brings unlimited text chats to free users