Gemma 4 Models - which vLLM version? Any PRs spotted?

[Digital_David]

@ben3241

  --default-chat-template-kwargs '{"enable_thinking": true}' \
  --chat-template examples/tool_chat_template_gemma4.jinja

I did see a reference to these flags in another post, but then are using llama-server with reasoning, Guide: Gemma 4 31B on DGX Spark via NemoClaw — Dual-Model Setup Guide - #2 by Digital_David

But when I try to use them with vLLM via spark-vllm-docker I run into an error. “Error: Missing parameter in recipe command: ‘“enable_thinking”’”

Is there some method needed to enable these or have the fixes been added int vLLM?