MiniMax H3 获 vLLM-Omni 首日支持，可通过兼容 OpenAI 的视频接口部署

logo logo # Videos API¶ vLLM-Omni provides an OpenAI-compatible video generation API for diffusion video models. The API supports asynchronous video jobs through `/v1/videos` and a synchronous benchmark-oriented endpoint through `/v1/videos/sync`. `/v1/videos` `/v1/videos/sync` Each server instance runs a single model specified at startup with `vllm serve <model> --omni`. `vllm serve <model> --omni` ## Quick Start¶ ### Start the Server¶ `vllm serve Wan-AI/Wan2.2-T2V-A14B-Diffusers --omni --port 8091` ### Create a Video Job¶ `create_response=$(curl -s \ -F "prompt=A cinematic tracking shot of a mountain lake at sunrise" \ -F "width=1280" \ -F "height=720" \ -F "num_frames=80" \ -F "fps=16" \ -F "num_inference_steps=40") video_id=$(echo "${create_response}" | jq -r '.id')` [...] `input_reference` `curl -s \ -F "prompt=continue this motion with consistent subjects and lighting" \ -F "[email protected];type=video/mp4" \ -F "width=1280" \ -F "height=720" \ -F "num_fra

vLLM-Omni is flexible and easy to use with: Heterogeneous pipeline abstraction to manage complex model workflows Seamless integration with popular Hugging Face models Tensor, pipeline, data and expert parallelism support for distributed inference Streaming outputs OpenAI-compatible API server Full-duplex realtime serving with streaming audio input and output (experimental) vLLM-Omni seamlessly supports most popular open-source models on HuggingFace, including: [...] Omni-modality models (e.g. Qwen3-Omni, MiniCPM-o 4.5, Cosmos3, HunyuanImage, BAGEL) TTS models (e.g. Qwen3-TTS, VoxCPM2, Ming-Omni-TTS, CosyVoice3) Diffusion models — image, video, and audio generation (e.g. MiniMax H3, Qwen-Image, Wan2.2, FLUX) Robot-policy and action models (e.g. GR00T-N1.7, DreamZero-DROID, InternVLA-A1, Cosmos3 action policy) ## Getting Started Visit our documentation to learn more. Installation Quickstart List of Supported Models Deployment Recipes for vLLM-Omni model serving ## Cont

vLLM-Omni is flexible and easy to use with: Heterogeneous pipeline abstraction to manage complex model workflows Seamless integration with popular Hugging Face models Tensor, pipeline, data and expert parallelism support for distributed inference Streaming outputs OpenAI-compatible API server vLLM-Omni seamlessly supports most popular open-source models on HuggingFace, including: Omni-modality models (e.g. Qwen3-Omni, Cosmos3, HunyuanImage, BAGEL) TTS models (e.g. Qwen3-TTS, VoxCPM2, Ming-Omni-TTS, CosyVoice3) Diffusion models — image, video, and audio generation (e.g. Qwen-Image, Wan2.2, FLUX) Robot-policy and action models (e.g. GR00T-N1.7, DreamZero-DROID, InternVLA-A1, Cosmos3 action policy) ## Getting Started Visit our documentation to learn more. [...] ## About vLLM was originally designed to support large language models for text-based autoregressive generation tasks. vLLM-Omni is a framework that extends its support for omni-modality model inference and serving:

The vLLM team just killed the last excuse for not serving omni-modal models in production. vLLM-Omni is live, 4.6k 🌟 stars in 8 months. Apache 2.0. And it's probably the most underrated open-source project in AI right now. Here's what most serving engineers are going to miss: Your current stack was built for one thing: text in, text out. Qwen-Omni, Gemini, GPT-4o, these models see, hear, and speak. Serving them with a text-only engine is like running a video call over SMS. vLLM-Omni closes that gap with three moves: 1️⃣ True omni-modality: text, image, video, and audio in a single inference pipeline. Not bolted on. Native. [...] The vLLM team just killed the last excuse for not serving omni-modal models in production. vLLM-Omni is live, 4.6k 🌟 stars in 8 months. Apache 2.0. And it's probably the most underrated open-source project in AI right now. Here's what most serving engineers are going to miss: Your current stack was built for one thing: text in, text out. Qwen-Omni, Gemini, GP

## Getting started with MiniMax H3 on Runpod The easy part: ComfyUI has day 0 support for Minimax and templates already set up. So you can get started today with just a few quick downloads. Here are the specs for sizing a GPU: First, deploy a pod using the official Runpod ComfyUI template with ~600 GB of volume disk; we’re going to pull the entire repo which will allow you to test each quant and decide what’s best for your use case. [...] ## Introducing MiniMax H3 Video generation has been the last major modality where the closed labs kept a comfortable moat. Weights stayed private, iteration was slow, and if you wanted frontier-quality video you rented it through someone else's API. That changed twice in the space of three days. MiniMax announced H3 on July 31, 2026 as an API-only product, then published the weights to Hugging Face on August 3 under the MiniMax H3 Community License. It's the strongest open-weight video model released to date, and unlike most of the "open" video r

Log inSign up ## Post user avatar MiniMax (official) @MiniMax\_AI Open weights are step one. Making them easy to serve is step two. H3 now has Day 0 support in @vllm\_project's vLLM-Omni, complete with an OpenAI-compatible video endpoint. Huge thanks to the vLLM team for bringing H3 into the open inference stack. ⚡ #MiniMaxH3 #OpenWeights user avatar vLLM @vllm\_project Aug 3 🎉 Congrats to @MiniMax\_AI on releasing the open weights for MiniMax H3! Day-0 support in vLLM-Omni! One model reads text, images, video, and audio as a single context and returns video with native stereo audio. Text-to-video, first/last-frame, and multi-reference generation, 4 00:00 3:06 AM · Aug 3, 202676.6KViews user avatar Emily @IamEmily2050 Aug 3 [...] Emily @IamEmily2050 Aug 3 Congratulations ❤️❤️❤️ 335 user avatar Knott Dynamics @KnottDynamics 19h Open weights are only the beginning. H3 makes quantized video generation practical on lo