Polymarket opened its GPT-5.6 release window this morning with more than $1.1 million staked across the contracts, and the tape itself tells the story: “Not released by June 28” leads at 58%, with “June 22–June 28” trailing at 40%. The market is hedging. ChatGPT Pro users, meanwhile, aren’t.

Over the weekend, subscribers traded screenshots and stopwatches showing that one-shot software builds GPT-5.5 used to finish inside ten minutes were now stretching past an hour when GPT-5.5 Pro was selected. The theory circulating on X is straightforward: OpenAI is A/B testing GPT-5.6 inside ChatGPT, swapped in silently for some Pro users. As of Sunday evening, OpenAI had confirmed nothing.

The supporting evidence has been arriving for two weeks. On June 10, The Information reported that OpenAI chief scientist Jakub Pachocki circulated an internal message describing GPT-5.6 as a “meaningful improvement” over GPT-5.5. AI influencer Leo, citing unnamed sources, says the model is “now being stealth tested when 5.5 Pro is selected in ChatGPT,” with a public launch planned for Thursday, June 25. A release candidate codenamed Kindle-Alpha has already leaked.

The commercial logic is the part the prediction market is underpricing. OpenAI filed a confidential S-1 with the SEC on May 22 and is targeting a public listing as early as September. A flagship model shipped into the road-show window isn’t coincidence; it’s narrative management. Per Tech Times, GPT-5.6 is the first OpenAI model trained with a redesigned pipeline that audits for cross-persona reward signal leakage before training begins, a direct response to the GPT-5.5 alignment failure.

The competitive frame sharpens the incentive. China’s GLM-5.2 already beats GPT-5.5 outright on FrontierSWE. Anthropic’s Mythos 5 and Fable 5 were blocked by a U.S. export-control directive issued June 12, removing the most credible Western challenger from several markets at exactly the moment OpenAI needs to look dominant to public-market investors.

Pachocki’s “meaningful improvement” reads less like a product claim than a pre-IPO disclosure rehearsal. The Polymarket bettors leaning toward “not this week” are betting against a company that has every reason to ship.

Sources