At a glance
MiniMax’s M3 announcement describes a model with a one-million-token context window and native multimodal support. Its API billing distinguishes requests at or below 512K input tokens from longer inputs.
Cost implication
A monthly token total cannot tell you how many requests cross that boundary. Standard and priority service levels also have different prices. Our calculator now offers an explicit request-size bracket for standard M3 pricing. Compare usage across different brackets in separate groups.
The announcement’s performance comparisons are vendor claims. This brief focuses on the documented billing distinction, rather than presenting those comparisons as our own tests.