hotai.news
← Back to AI Newsflash
AI News / Event details
vllm.ai · Published

vLLM-Omni has enabled Day-0 support for MiniMax H3-Image on Vera Rubin, achieving a median latency of 0.87 seconds for 4K generation.

Automated summary · Based on collected summaries; not individually verified
Prepared by HotAI

What happened

The vLLM-Omni team has ported the MiniMax H3-Image open-weight image generation and editing model to NVIDIA's Vera Rubin supercomputer. With a four-card SP4 configuration, it achieves native 4096×4096 generation with a median latency of 0.866 seconds—2.21 times faster than the GB200—and delivers a throughput of 6,115 images per hour using an SP1×DP4 setup.

Background

What is MiniMax H3? Multimodal video generation and its relationship to Hailuo

MiniMax H3 is a multimodal generation system that interprets text and audiovisual references and produces video with native stereo sound. It follows the development of Hailuo 01 and 02 but is not simply a new alias for them. Hailuo AI is a creative product interface; H3 is a model system accessible through relevant services or open weights.

What Is MiniMax? Language, Speech and Video Model Lines

MiniMax is an AI company and model brand. Its models cover language and coding, with separate speech, music and video generation lines. “Using MiniMax to code” generally requires a specific M-series language model, which is distinct from the MiniMax Agent app or a Speech model. The exact model determines input types and thinking controls, while an older weight license does not replace current hosted service terms.

Reading notes

The event summary comes from collected excerpts. Background passages quote existing HotAI articles to explain terms and do not establish new facts about this event. Check the original sources for product results, conditions and subsequent updates.

Original source

Sources and collection timeline

SourceReport titleOriginal publication timeCollected by HotAI
vllm.aivLLM-Omni has enabled Day-0 support for MiniMax H3-Image on Vera Rubin, achieving a median latency of 0.87 seconds for 4K generation.2026-11-12 08:002026-11-12 16:57