MiniMax · Models

MiniMax M2.7 on tap.

MiniMax M2.7 brings long-context reasoning and competitive performance to the gateway — a strong, economical pick for agents and high-volume work.

Long context

Keeps large documents and conversations in view for coherent output.

Strong reasoning

Capable multi-step reasoning and tool use for agent workloads.

High volume value

Competitive pricing routed at the provider rate, zero markup.

Available on the gateway

minimax/minimax-m2.7Flagship long-context
minimax/minimax-m2.7-liteFaster, economical

All models route through the MegaBrain Gateway at the provider rate — 0% inference markup. Switch between them in one click.