Expired50% confidenceFactExact time
MiniMax M3 can directly understand long video content, comprehending a 12-minute NVIDIA launch event video and generating an HTML-format news brief, by directly 'watching' the video itself rather than relying on speech-to-text conversion.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/2/2026
First Seen
This claim has expired. It is kept for the record and no longer describes the current state.
Valid until: 9/30/2026(expired)
Sources
Related Claims
UnverifiedMiniMax M3 features a million-token-level ultra-long context window combined with audio and video multimodal understanding capabilities.76% similarVerifiedMiniMax M3 is a multimodal model that supports text, video, images, and other modalities74% similarUnverifiedThe core strengths of the MiniMax M3 model are its ultra-long context processing capability and multimodal comprehension capability (capable of handling text, images, and video).73% 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% similarVerifiedMiniMax H3 is a multimodal video generation model capable of simultaneously generating video frames and matching audio (including dialogue, sound effects, and ambient sound).71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/49537API
curl https://kongchang.com/api/v1/knowledge/claims/49537MCP
get_claim(id=49537)