OpenAI Retires GPT-4o, GPT-5; Rolls Out GPT-5.3 and GPT-5.5 in ChatGPT
The AMW Read
Retiring GPT-4o and GPT-5 within a year meaningfully updates the segment's model lifecycle pattern, but does not introduce a new entrant or resolve a debate; significance is segment-level due to accelerated churn and tiered access.
OpenAI Retires GPT-4o, GPT-5; Rolls Out GPT-5.3 and GPT-5.5 in ChatGPT
OpenAI has retired GPT-4o, GPT-4.1, GPT-4.1 mini, o4-mini, and GPT-5 from ChatGPT as of February 13, 2026, replacing them with GPT-5.3 as the default model and GPT-5.5 (Thinking and Pro) for advanced reasoning. GPT-5.5 Pro is available only to Pro, Business, and Enterprise tiers. Free users retain GPT-5.3 with a 10-message limit every 5 hours; Plus users get 160 messages per 3 hours. API access for the retired models remains unchanged. GPT-4o will persist in Custom GPTs for Business/Enterprise/Edu until April 3, 2026.
This rapid model churn—retiring multiple flagship models barely a year after GPT-5 launched—signals an accelerating capital-compression arc in foundation models. OpenAI is aggressively amortizing R&D by retiring older inference cost centers and pushing users toward the latest architecture, which likely benefits from efficiency gains and lower per-token costs. The tiered availability (Pro for highest capability, limits for free/Plus) reflects a hyperscaler-distribution moat where access and user experience are dynamically gated by willingness to pay. By making GPT-5.5 Thinking the default reasoning engine for complex tasks, OpenAI is also training user expectations: instant answers for easy questions, deeper reasoning for hard work, with the model routing transparently. This bifurcation may compress margins for standalone reasoning-model competitors who lack a captive consumer distribution channel.
The strategic implication is clear: OpenAI is racing up the capability curve while simultaneously commoditizing its older models, forcing the rest of the ecosystem to either keep pace on performance or find a differentiated niche. Enterprise customers on Business/Enterprise plans get unlimited access, reinforcing the pattern that the real monetization lies in high-volume, high-velocity use cases within organizations that can absorb the subscription cost.
#OpenAI #GPT5 #FoundationModels #ModelRetirement #AICompetition #HyperscalerDistribution
