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

2026-04-09 06:52:48

Abstract generation in progress

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’s view is that its advantage lies in a cost reduction of about 20x and a speed increase of 2–4x. Taking an example of outputting 10 million tokens per day, the cost of using Opus 4.6 is about $250/day, while MiniMax M2.7 requires only about $12/day. This progress is driven by open benchmarks such as SWE-Rebench and Terminal Bench 2.0, making open models a feasible choice for deploying agents in production environments, and they can also be used in conjunction with closed-source models. (Source: InFoQ)

View Original

This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.