MiniMax: MiniMax M2

MiniMax M2 is a text-in, text-out model from MiniMax with a 204,800-token context window and a near-matching 196,608-token output limit, which makes it practical for long-document work. It supports tool use and reasoning, giving it coverage for agentic and multi-step tasks. Structured output support is unconfirmed from available specifications. At $0.255 per million input tokens and $1.00 per million output tokens, it sits in the budget tier for reasoning-capable models. Its blended benchmark score of 46.3 is drawn from a single independent benchmark, so treat that figure as a preliminary signal rather than a settled verdict. Teams looking for an inexpensive option with a large context window and tool-calling support may find it worth testing, but those who need strong benchmark validation before committing should wait for broader evaluation coverage to emerge.