Unverified50% confidenceFactExact time
MiniMax H3在图生视频任务中通过时序建模和身份保持模块，能完整保留原图中的人物形象与画风特征，不出现形象崩坏
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
8/26/2026
First Seen
Valid until: 11/24/2026
Sources
Related Entities
Related Claims
VerifiedMiniMax-H3 features two capabilities: image-to-video generation and reference-based generation.79% similarUnverifiedAfter integration into ComfyUI, MiniMax H3 supports six core workflows: text-to-video, image-to-video, first-and-last-frame animation, reference-image-to-video, audio synchronization, and image editing.74% similarUnverifiedMiniMax previewed in a Reddit AMA that its new solution based on the H3 architecture will have a single model handle both text-to-image generation and image editing, with plans to open-source the full-weight image model.72% similarUnverifiedBecause H3 is essentially a video model rather than an image model, it falls short in the detail quality of static images.72% similarUnverified借助 MiniMax H3 模型与 ComfyUI 上下文循环节点，可以在本地设备上生成时长超过一分钟的长视频72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/803429API
curl https://kongchang.com/api/v1/knowledge/claims/803429MCP
get_claim(id=803429)