MiniMax AI has announced that its M2.7 model, when operating at 400 transactions per second, achieves nearly imperceptible latency. This performance milestone was achieved with the assistance of the SambaNova Systems team. AI
Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →
IMPACT Achieving near-imperceptible latency at high transaction rates could improve user experience for real-time AI applications.
RANK_REASON This is a performance update on an existing model, not a new release or significant industry event.