vLLM v0.31.0rc3 Adds Randomized Dummy Input Support for Model Runner V2

Core summary

One sentence to understand this update

vLLM's v0.31.0rc3 release introduces support for randomized dummy inputs within its Model Runner V2, with Claude Opus 5.5 listed as a co-author.

Impact & opportunity

What this could mean

This feature can aid developers in testing and benchmarking model performance more flexibly, potentially streamlining development workflows for vLLM users.