MiniMax M2.5 is a large language model (LLM) designed to enhance real-world productivity, particularly in coding and agentic workflows. MiniMax, a Shanghai-based AI company founded in early 2022, created M2.5. The company aims to advance AI towards AGI (Artificial General Intelligence).
M2.5 excels in coding, demonstrating proficiency across multiple programming languages and handling the entire development lifecycle, from system design to testing. It exhibits "spec behavior," planning architecture and UI design before writing code. MiniMax M2.5 has shown significant improvements in advanced workspace scenarios like Word, PPT, and Excel. It is designed for high-throughput, low-latency environments. M2.5 also supports multilingual coding tasks. There are two versions of the model, M2.5 and M2.5-Lightning, which have identical capabilities but different speeds.
MiniMax M2.5 is available through the MiniMax Open API Platform. Pricing is offered through a "Coding Plan" or a "Pay-As-You-Go" option. MiniMax M2.5 has a cost of $0.30 per 1 million input tokens and $1.20 per 1 million output tokens. The faster M2.5-Lightning model costs $0.30 per million input tokens and $2.40 per million output tokens. According to news reports from February 2026, M2.5 has just been released and is currently available.
