Skip to content
News Product launch

OpenAI simplified ChatGPT's model picker on June 10, replacing Thinking tiers with Instant, Medium, High, and Extra High

· by Pondero Newsdesk

The short version

OpenAI relabeled the ChatGPT model picker on June 10, 2026, dropping Thinking Light and renaming three remaining thinking tiers as Medium, High, and Extra High. Plus and Pro users can now also set Instant to auto-promote to Medium for demanding tasks.

OpenAI simplified ChatGPT's model picker on June 10, replacing Thinking tiers with Instant, Medium, High, and Extra High

OpenAI updated the ChatGPT model picker on June 10, 2026, swapping out the Thinking-tier naming for four plain-language options and removing one tier entirely. The change rolled out to Plus and Pro users on web, iOS, and Android globally that same day.

What changed

Per OpenAI's official ChatGPT release notes, the picker now lists six options: Instant, Medium, High, Extra High (Pro plans only), Pro Standard (Pro plans only), and Pro Extended (Pro plans only). The underlying GPT-5.5 model family powers all of them, per the changelog.

The renaming maps directly from the old Thinking-tier labels. Thinking Standard became Medium, Thinking Extended became High, and Thinking Heavy became Extra High. Instant kept its name. Thinking Light was removed without a replacement. The two Pro-exclusive tiers, Pro Standard and Pro Extended, were not renamed.

An auto-switch setting is new. Users can allow Instant to promote itself to Medium when ChatGPT judges a task requires more reasoning effort. The toggle lives under General > Settings, and it is opt-in.

The picker appears at the top of the conversation on iOS and Android and in the message composer on web.

Why it matters

The Thinking-tier naming caused friction because it implied a binary between "fast but shallow" and "slow and thoughtful" without giving users a feel for the steps in between. Four tiers with everyday words (Instant, Medium, High, Extra High) communicate a cost-quality spectrum in terms most users understand without needing to read documentation.

Removing Thinking Light is the more substantive cut. It occupied the slot between Instant and Thinking Standard, and its removal leaves the Instant-to-Medium auto-switch as the only path for tasks that need slightly more effort than the fastest tier can deliver. Whether the auto-switch produces equivalent output to what Thinking Light covered is a practical question OpenAI did not address in the changelog.

The naming alignment also matters for how users discuss and share ChatGPT settings. Support threads, tutorials, and screenshot-based help content had already forked between users on old Thinking labels and users who saw earlier partial rollouts. The unified naming should close that gap.

What to watch next

Three things are worth tracking. First, whether the Instant-to-Medium auto-switch becomes a source of unexpected token usage for users who leave it on by default. Second, whether the Extra High tier drives measurable conversion to Pro plans. Third, whether a parallel simplification reaches the API's reasoning_effort parameter, which still uses its own naming scheme. OpenAI has not announced API-side changes tied to this update.

Sources