Data compiled by Woofun AI shows that the probability of GPT-5.6 failing to reach public availability before June 28 has climbed to 78%, marking a 67% increase within a 24-hour window. This surge in betting volume follows an internal memo dated June 11 from OpenAI Chief Scientist Jakub Pachocki, who characterized GPT-5.6 as a substantial upgrade over its predecessor, GPT-5.5.

Analysts attribute the heightened market focus to earlier leaks regarding the Codex backend in May and the accelerated development cycle observed after the April 23 release of GPT-5.5. The Polymarket event strictly defines eligible models as those explicitly named GPT-5.6 or identified by OpenAI as direct successors, such as GPT-5.7 or GPT-5.8. Task-specific variants, cost-optimized releases like Nano or Mini, o-series reasoning models, and next-generation flagships like GPT-6 are excluded from settlement criteria. Resolution will depend on official OpenAI announcements confirming open beta or public waitlist access, with closed beta or private access versions deemed ineligible.