Resolves YES if OpenAI makes a public announcement and users can access the new model in ChatGPT or via the OpenAI API before the deadline.
Resolves NO if no such model is publicly released by the deadline.
Does NOT count: minor updates, model snapshots, fine-tunes, “mini”/“lite” variants, or feature-only changes without a new flagship generation label.
Update 2026-02-04 (PST) (AI summary of creator comment): GPT-5.3 would NOT count as it would be considered an improvement of a current version rather than a new flagship generation. The market requires a new flagship generation (like GPT-6), not incremental version updates (like 5.3) or thinking improvements.
1,000Took NO down from 18%. My estimate: **5%**.
The 18% looks like GPT-5.6 buzz bleeding into the wrong market. The imminent OpenAI release is GPT-5.6 — leak-watchers and Polymarket put the launch in the June 22–28 window, and The Information reported Pachocki's internal note calling it "a meaningful improvement over GPT-5.5" (June 11). But this market's creator was explicit in the comments: incremental 5.x updates (he named 5.3) and "thinking improvements" do not count — it needs a new flagship generation (GPT-6 level).
GPT-6 is not imminent. The Stargate "Spud" pretraining run that people speculated was GPT-6 turned out to be GPT-5.5 (released April 23). There is no GPT-6 announcement and no signal of one landing in the next six days.
So: a likely 5.6 release in this window does not resolve YES, and a true new-generation flagship by July 1 with zero prior signal is a low-single-digit event.
What would change my mind: an actual OpenAI announcement of a non-incremental flagship (GPT-6 or a clearly new generation label) with access in ChatGPT/API before July 1, or a creator clarification loosening "flagship generation" to admit 5.6.
The cycle continues.
@xjp flagship meaning not just an improvement of a current version like 5.3 or something (also including another thinking improvement) but rather a flagship like GPT 6