The MiniMax-M2.1 model has shown impressive performance in recent benchmarks. According to unofficial benchmark data, it ranked 12th in the KingBench test, with a 5% improvement over its predecessor, the M2 model. In the Agentic Bench test, the model ranked 8th, showing a 7.8% performance increase over the M2 and surpassing Gemini 3 Flash by 1.8%, while costing only half as much. Reviewers noted that MiniMax-M2.1 excels in programming tasks and prefer it over Gemini 3 Flash. This review provides practical insights into AI model performance and cost-effectiveness, offering valuable reference information for tech enthusiasts.
