3OpenAI Release Notes·
OpenAI restores Extended thinking level for GPT-5.2 Thinking in ChatGPT after inadvertent reduction
OpenAI disclosed a series of adjustments to thinking time settings for GPT-5.2 Thinking in ChatGPT, including an unintentional reduction of the Extended thinking level in January 2026 that has now been corrected. Standard and Light thinking times were also deliberately reduced based on user preference data favoring faster responses. The update clarifies that thinking time is tuned independently per model and is not cross-model comparable, and that a thinking level toggle introduced in September 2025 gives users explicit control over the tradeoff between speed and reasoning depth.
Related guides (4)
Related events (8)
OpenAI rolls out GPT-5.4 mini in ChatGPT as reasoning fallback
OpenAI is deploying GPT-5.4 mini in ChatGPT, making it available to Free and Go users via the 'Thinking' feature and as a rate-limit fallback for GPT-5.4 Thinking for paid users. The model will not appear as a selectable option in the model picker, positioning it as a background capacity-management tool rather than a user-facing choice. Enterprise customers retain the option to set Auto routing to default to GPT-5.4 mini.
OpenAI retires GPT-5 Instant and Thinking variants from ChatGPT
OpenAI retired the GPT-5 Instant and Thinking model variants from ChatGPT on February 13, 2026. The retirement affects only the ChatGPT product; no API changes are included. This signals a product lifecycle transition as newer flagship models supersede these variants.
OpenAI releases GPT-5.4 with integrated reasoning, coding, and agentic capabilities in ChatGPT
OpenAI released GPT-5.4, a frontier model that consolidates recent advances in reasoning, coding, and agentic workflows, incorporating the coding capabilities of GPT-5.3-Codex. The ChatGPT deployment introduces a 'Thinking' mode that surfaces an upfront reasoning plan mid-response, allowing users to redirect the model before it completes. Additional improvements include enhanced deep web research for specific queries, better long-context retention during extended thinking, and improved context window management for faster, higher-quality outputs.
OpenAI retires GPT-5.2 models from ChatGPT, migrating users to GPT-5.5
As of June 12, 2026, OpenAI has removed GPT-5.2 Instant, GPT-5.2 Thinking, and GPT-5.2 Pro from ChatGPT, with existing conversations automatically continuing on corresponding GPT-5.5 models. The retirement follows OpenAI's stated policy of keeping models available for approximately 90 days after a successor is released. This reflects OpenAI's ongoing model lifecycle management as the GPT-5.x generation matures.
OpenAI retires GPT-5.1 models from ChatGPT, migrates users to GPT-5.3/5.4
OpenAI removed GPT-5.1 Instant, GPT-5.1 Thinking, and GPT-5.1 Pro from ChatGPT as of March 11, 2026. Existing conversations using GPT-5.1 variants are automatically migrated to corresponding current models: GPT-5.3 Instant, GPT-5.4 Thinking, and GPT-5.4 Pro. This is a routine model lifecycle action reflecting OpenAI's cadence of deprecating older model versions as newer ones become the standard.
OpenAI simplifies ChatGPT model picker for Enterprise, Edu, and Business tiers
OpenAI is rolling out a redesigned model picker for ChatGPT Enterprise, Edu, and Business across web, iOS, and Android that replaces reasoning-tier labels with plain-language speed/effort options: Instant, Medium, High, Extra High, Pro Standard, and Pro Extended. The rename maps Medium→Thinking Standard, High→Thinking Extended, Extra High→Thinking Heavy, while Thinking Light is discontinued. Pro Standard and Pro Extended continue to use GPT-5.5 Pro. Access controls remain tied to workspace-level Reasoning Model Access and Pro Model Access settings.
OpenAI deploys GPT-5.5 Instant to improve ChatGPT health and wellness responses
OpenAI has updated ChatGPT's health and wellness capabilities using GPT-5.5 Instant, citing improvements in reasoning, contextual understanding, and communication clarity. The update was informed by physician evaluations, suggesting a structured clinical validation process. This represents both a model deployment signal and a domain-specific capability push into health intelligence.
OpenAI updates GPT-5.5 Instant with improved response style, removes canvas support
OpenAI is rolling out a quality update to GPT-5.5 Instant in ChatGPT and the API, targeting response style improvements including more natural conversational tone, better pacing, and fewer overly long or bullet-heavy outputs. The update also removes canvas from GPT-5.5 Instant and GPT-5.5 Thinking, replacing that functionality with inline writing blocks and code blocks in chat responses. Paid users retain canvas access via legacy models until those are sunset.