Ramine Roane says AMD is currently the leading company contributing code to vLLM, with 502 core commits in 90 days, 13% of all organization contributions, and nearly four times Nvidia. He also credited Red Hat, IBM, Embedded LLM, and Inferact.
Google Gemma says seeding a canvas with a response template in vLLM lets DiffusionGemma return confidence scores and probability distributions for yes/no, multiple-choice, and scored questions in a single denoising step.
Contributors from IBM, Meta, and Hugging Face present hardware-agnostic layers intended to balance frontier performance with portability for users running diverse models on diverse hardware.
vLLM says it has integrated PyNvVideoCodec to offload video decoding from CPU to GPU NVDEC, reporting 2x+ throughput at 8×H100. The project says the change ships with CUDA vLLM releases.
vLLM said Qwen-Image-2.1 from @Alibaba_Qwen has day-0 support in vLLM-Omni, describing a 7.1B DiT paired with Qwen3-VL-8B for generation, editing, and transparent images. Qwen thanked vLLM and said the one model is ready to serve.
