Independent assessments show that MiniMax M2.7 rivals leading closed-source models in core intelligent agent tasks.

ME News message, April 5 (UTC+8). Recently, according to LangChain’s independent evaluation, the open-weight model MiniMax M2.7 has reached a level comparable to that of closed-source frontier models on core agent tasks such as file operations, tool calling, and instruction following. The article argues that its advantage is a substantial cost reduction of about 20 times, with speed improved by 2–4 times. Taking an example of 10 million tokens output per day, the cost of using Opus 4.6 is about $250/day, while MiniMax M2.7 only requires about $12/day. This progress has been driven by open benchmarks such as SWE-Rebench and Terminal Bench 2.0, making open models a viable option for deploying agents in production environments, and they can also be used in combination with closed-source models. (Source: InFoQ)