Hi, there. I want to know whether I can run vllm in Docker to serve Wan2.2 Text-To-Video Model？
I’m considering to buy MacBookPro with M4 Pro/M5.
Any suggestion is appreciated 
It seems likely that vLLM does not currently support running on MacBook Pro with M4/M5 (Apple Silicon), as both the official documentation and user reports indicate that vLLM is not supported on Apple Silicon/MacOS, even in Docker. There is no mention in the data sources of support for the Wan2.2 Text-To-Video model specifically, and vLLM’s multimodal support is focused on certain vision and video models, but not on Mac hardware or with Apple GPUs/CPUs. Running vLLM in Docker on MacBook Pro (M4/M5) is not supported at this time, and you may encounter compatibility issues or lack of hardware acceleration for LLM inference and serving on Apple Silicon devices. For text-to-video models, there is no explicit support or example for Wan2.2 in the vLLM documentation.
Would you like more detail on supported hardware or alternative deployment options for text-to-video models?
Sources:
Hint: Mention @RunLLM in the post for followups.