The a16z Show · Thursday, August 6, 2026
Simon Mo states that VLLM supports over a thousand model architectures, including many open-weight models, often seeing new models live on VLLM immediately upon release. He also mentions that VLLM works closely with hardware vendors like Nvidia, AMD, and Google, frequently serving as a benchmark for their newest chips.
“So that means across like Nvidia, AMD, Google and Amazon, and a lot more, their newest chip will make sure VLM can run on them.”
“And then in many cases they use VLM as a benchmark to make sure it runs well on them.”
“Because for VLM, we support more than a thousand model architectures up to today, and a lot of those are proprietary, but also a lot of those are open weight.”