The MiniMax Hailuo 3 Max model has demonstrated a generation time of approximately 10 seconds per sequence. This performance was observed in a test where the model generated short image-to-video sequences, which were then compiled together. The model is still in a locked state, with the first frame generation taking this duration. AI
IMPACT Demonstrates a specific performance metric for an image-to-video generation model, potentially indicating progress in generation speed.
RANK_REASON The item reports on a specific performance metric (generation time) for an AI model, which falls under research milestones. [lever_c_demoted from research: ic=1 ai=1.0]
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →