MiniMax posted a short teaser on June 1 confirming their next model is around 10 days out - putting the release somewhere around June 11.
That's the full announcement. No benchmark numbers, no parameter count, no word on whether this will ship as an open-weight model (downloadable and self-hostable) or stay API-only like some of their recent releases. MiniMax-01, their last major release, was a 456-billion-parameter mixture-of-experts model that turned heads for its 1-million token context window - roughly enough to fit 2,500 pages of text - and competitive performance against much larger closed models.
The Chinese AI lab has been quietly consistent, shipping capable models at a cadence that outpaces many Western competitors. Whether this next release is an incremental update or something more substantial, we'll know shortly. No action needed now - just worth knowing something is coming before the week of June 9.