MiniMax H3

MiniMax H3 is an advanced multimodal AI video generation model that creates 4–15 second videos from text prompts, reference images, video clips, and audio files, with output resolution up to 2K.

Key features include:

• Text-to-Video: Generate videos from detailed natural language prompts describing subject, action, setting, lighting, and camera movement.
• First & Last Frame Control: Upload starting and/or ending frame images to guide video transitions and animation.
• Multimodal References: Combine up to 9 images, 3 videos (15 seconds total), and 3 audio files (15 seconds total) to provide precise creative direction for subject, motion, camera, and visual style.
• Flexible Output: Choose 768P or 2K resolution, durations from 4 to 15 seconds, and multiple aspect ratios (21:9, 16:9, 4:3, 1:1, 3:4, 9:16).
• Motion & Camera Guidance: Describe pans, zooms, tracking shots, and static shots directly in prompts.

Target audiences include marketing & product teams, ecommerce businesses, social media creators, filmmakers, game developers, and motion designers. Use cases span product ads, social videos, film previsualization, game concept development, and motion design exploration.

MiniMax H3 operates on a credit-based pricing model with one-time purchase packs (Starter, Creator, Studio, Business) – credits never expire and downloads are watermark-free. New users receive 50 starter credits to try the service.