GPT-6 'Spud' April 14 Launch Debunked, Release Expected Later This Spring
The viral April 14 launch date was a bust, but prediction markets give 78% odds for a release by April 30.
The highly anticipated April 14, 2026 launch of OpenAI's next-generation AI model, codenamed 'Spud,' turned out to be a false rumor. Despite widespread speculation and cryptic teases from OpenAI employees about 'next week,' no official announcement, blog post, or product launch occurred. The rumor originated from an unverified leak on April 7, which claimed a 40% performance improvement over GPT-5.4 and a 2M-token context window alongside a unified super-app. OpenAI's only official confirmations are that pretraining finished on March 24 and CEO Sam Altman stated a launch is 'a few weeks' away.
While the specific date was wrong, the broader launch window remains intact. Community prediction markets, often a reliable leading indicator, currently give approximately 78% odds that 'Spud' will launch by April 30, 2026, and 95% odds by June 30. The model's final branding is still undecided, with internal signals pointing both to 'GPT-5.5' and 'GPT-6.' CTO Greg Brockman has positioned it as 'two years of research' and 'not an incremental improvement,' suggesting a major leap. The launch is expected to be tied to a new, unified product ecosystem combining ChatGPT, Codex, and the Atlas browser.
- The viral April 14 launch date was debunked; no official announcement was made.
- Prediction markets give ~78% odds for a launch by April 30, 2026, based on insider teases and confirmed pretraining completion.
- The model's name (GPT-6 vs. GPT-5.5) is undecided, but it's confirmed to be a major, non-incremental release after two years of R&D.
Why It Matters
The next OpenAI model will set a new benchmark for AI capabilities, impacting developers, businesses, and the competitive landscape.