MiniMax M3.1 Flash Released with High Speed and Low Cost
MiniMax has officially released the M3.1 Flash model, revealing that the stealth model 'Space Bunny' tested on OpenRouter in late September was an earlier version. The company announced plans to release the model weights on October 23. The model offers significantly higher efficiency, delivering three times the speed at one-third the cost of the previous M3 version. This combination of affordability, high performance, and an upcoming open-weights release makes it highly attractive for developers and the local LLM community. MiniMax M3.1 Flash features a 1-million-token context window with configurable thinking effort levels and API pricing set at $0.15 per million input tokens and $0.60 per million output tokens. Its reported benchmark scores include 62% on Automation-Bench, 30% on ALE, 74% on DeepSWE, and 91% on BabyVision, though the exact parameter count has not been disclosed yet.
## BACKGROUND
OpenRouter is a unified platform layer that allows developers to access hundreds of different AI models through a single API interface. Context windows refer to the maximum amount of text data a model can process at once, while open weights allow developers to download and run models locally on their own hardware.