MiniMax M3.1 Flash Enables Higgsfield Creator to Weave AI-Generated Knitwear Narrative

MiniMax's new M3.1 Flash multimodal model is fueling creative experiments on Higgsfield, the AI-native video platform, with early projects demonstrating how frontier reasoning and native video understanding can drive ambitious narrative work. Shortly after MiniMax released M3.1 Flash on October 19, 2026, independent creators began exploring the model's multimodal capabilities through Higgsfield's integrated workflow. The model, trained on over 200 trillion tokens and combining text, image, and video inputs natively, handles both complex reasoning (with five adjustable effort levels) and visual generation in a single pass—a significant departure from traditional multi-step creative pipelines.