MiniMax: MiniMax M2.1

MiniMax M2.1 is a text-only model from MiniMax with a 204,800-token context window and a maximum output of roughly 196,600 tokens, making it practical for long-document work. It supports tool use and reasoning, which covers most agentic and multi-step task patterns. Structured output support is unconfirmed from available data, so verify that before committing to schema-dependent pipelines. At $0.29 per million input tokens and $0.95 per million output tokens, MiniMax M2.1 sits at a low price tier, which makes cost a reasonable argument for it. However, its blended benchmark score of 51.4 is drawn from only one independent benchmark, so its quality profile is largely unproven against the broader model field. Teams with tight budgets and workloads that lean on long context or reasoning may find it worth testing, but buyers who need strong evidence of benchmark reliability should wait for wider coverage.