OpenAI's GPT-6 (Spud) Launch Delayed Beyond April 14, Anticipated Soon

The anticipated April 14 launch of OpenAI's next-generation AI model, internally codenamed 'Spud,' did not materialize, leaving the tech community waiting. While OpenAI has confirmed pretraining finished on March 24, 2026, and executives have teased a launch 'a few weeks' away, the company has provided no official date, model card, or API details. The April 14 rumor stemmed from an unverified leak and cryptic employee tweets, but lacked corroboration from major tech press. Prediction markets like Polymarket currently assign a 78% probability that the model—whether branded as GPT-6 or GPT-5.5—will launch by April 30, with odds rising to 95% by June 30.

This delay has created a strategic opening for competitor Anthropic, which seized the moment to ship concrete product updates. Anthropic released Claude Code Routines and a desktop app redesign, with reports from The Information indicating its next major model, Opus 4.7, and a new AI design tool could launch 'as soon as this week.' The naming of OpenAI's model remains undecided, with insiders leaning toward 'GPT-5.5,' though CTO Greg Brockman has positioned 'Spud' as a non-incremental improvement representing 'two years of research.' The launch is now expected in a late April to late May window, shifting the competitive narrative from a solo OpenAI reveal to a potential head-to-head release clash with Anthropic's next wave.