MiniMax-M3.1-Flash-Preview goes live with a 1M context window, available only through the Token Plan and MiniMax Code
As of September 27, 2026, MiniMax M3.1 Flash Preview is described as a frontier multimodal coding model with a 1M context window and tunable thinking depth, an…
Published on MyPrivateClaw
Sep 29, 2026, 1:18 AM UTC
Coverage date
Sep 28, 2026
Last updated
Sep 29, 2026, 1:18 AM UTC
News summary
MiniMax M3.1 Flash Preview is a frontier multimodal coding model with a 1M context window and tunable thinking depth. As of September 27, 2026, it is available only through the Token Plan and MiniMax Code rather than a broadly documented public API, with distribution limited to MiniMax Code and Token Plan for the launch window. No full public parameter card or open API sheet for the Preview model appears in reviewed sources. Why it matters Launch window use is confined to MiniMax Code and the Token Plan, so access runs through those two surfaces. Because no broadly documented public API was published for the Preview model in reviewed sources, other documented integration routes are not described. The stated 1M context window and tunable thinking depth therefore sit behind those surfaces as of September 27, 2026.