MiniMax M3.1 Flash goes free on OpenRouter

MiniMax M3.1 Flash, a 1M-context multimodal coding model, is now available for free via OpenRouter. It was previously limited to MiniMax Code and the Token Plan subscription. The model targets everyday coding and always runs with reasoning enabled at a selectable effort level.

• 1M-token context with text, image and video input and text output
• MoE with roughly 428B total and about 23B active parameters per token
• Reasoning cannot be disabled; effort levels are low, medium, high, xhigh and max
• Early users report 90–110 tokens/sec decode, with no official benchmarks yet