Chinese artificial intelligence startup MiniMax has introduced an updated model named M3.1 Flash. According to the company, the release introduces significant advancements in coding, cybersecurity, and multi-step AI tool workflows. MiniMax additionally reports that the model delivers nearly triple the response speed of previous versions, accompanied by a 70% reduction in usage costs. This rollout intensifies the ongoing competition among AI developers to deliver faster and more affordable tools.
MiniMax states that M3.1 Flash assists with writing code, resolving software bugs, operating computer tools, and managing tasks across productivity applications. The company also highlights strong performance outcomes in its internal evaluations. Nevertheless, these assertions still require validation through independent testing and real-world deployment. The model is accessible via MiniMax’s API, M Plan, and MiniMax Code, and the firm intends to release the model files at a later date so developers can access them outside of existing platform services.
Writing software code is often time-consuming, particularly during debugging and error correction. A quicker AI model could assist programmers in completing these assignments more rapidly while managing extended multi-step processes with minimal human intervention, thereby reducing routine workloads for development teams. Even so, velocity is only one component; the model must also deliver accurate answers and functional code.
Also Read: Microsoft Brings AI Agents to PCs with New Coding, Security Tools
Lower AI Costs Could Help Startups and Big Firms
AI models consume computing resources every time they process a prompt and generate a response, an expense known as inference cost. MiniMax prices M3.1 Flash at USD 0.15 per million input tokens and USD 0.60 per million output tokens, with tokens representing the fundamental units of text that the system reads or generates.
Combined, the listed price for one million input tokens and one million output tokens totals USD 0.75. In comparison, the same workload using MiniMax’s M3 model would cost approximately USD 1.50, illustrating how reduced pricing can benefit organizations that submit high volumes of requests.
Startups stand to gain significantly if the model performs as advertised, provided they verify its quality and speed before full integration. While M3.1 Flash presents compelling performance and pricing claims, practical application will ultimately determine its actual value.
