Next OpenAI Public Model Release
Method: Range and median of the last 4 public release gaps (56d, 19d, 7d, 27d): shortest gap 7 days (2026-11-02), median gap 23 days (2026-11-18), longest gap 56 days (2026-12-21).
PUBLIC ACCOUNTABILITY LEDGER · DATA/FORECASTS.JSON
Anyone can make vague predictions and quietly edit them later. Every release window on ModelRumors.ai is committed to
data/forecasts.json with the date it was made, the start and end of the window, and the plain-language gap calculation. When a lab ships, we record the actual release date and whether it landed inside the window.
ALL LOGGED FORECASTS (6)
Read hits with the window width in mind. A window spans the shortest to the longest of the lab's last four release gaps, so a lab with uneven gaps gets a wide window that is easy to land in. Current windows average 48 days wide. A narrow window that hits says more than a wide one.
Method: Range and median of the last 4 public release gaps (56d, 19d, 7d, 27d): shortest gap 7 days (2026-11-02), median gap 23 days (2026-11-18), longest gap 56 days (2026-12-21).
Method: Range and median of the last 4 public release gaps (39d, 21d, 6d, 9d): shortest gap 6 days (2026-10-13), median gap 15 days (2026-10-22), longest gap 39 days (2026-11-15).
Method: Range and median of the last 4 public release gaps (63d, 23d, 20d, 41d): shortest gap 20 days (2026-11-02), median gap 32 days (2026-11-14), longest gap 63 days (2026-12-15).
Method: Range and median of the last 4 public release gaps (51d, 69d, 35d, 40d): shortest gap 35 days (2026-10-26), median gap 46 days (2026-11-06), longest gap 69 days (2026-11-29).
Method: Range and median of the last 4 public release gaps (98d, 13d, 28d, 37d): shortest gap 13 days (2026-10-30), median gap 32 days (2026-11-18), longest gap 98 days (2027-01-23).
Method: Range and median of the last 4 public release gaps (5d, 23d, 40d, 3d): shortest gap 3 days (2026-10-18), median gap 14 days (2026-10-29), longest gap 40 days (2026-11-24).