[Bug] Qwen reasoning parser

[ahao-anyscale]

Purpose

When running with Qwen/Qwen3-30B-A3B, reasoning_content is empty despite correctly formed <think> tag. Relevant issue: ray-project/ray#57919

Test Plan

Run requests with Qwen/Qwen3-30B-A3B and observe that reasoning_content is populated instead of content.

---

Essential Elements of an Effective PR Description Checklist

• The purpose of the PR, such as "Fix some issue (link existing issues this PR will resolve)".
• The test plan, such as providing test command.
• The test results, such as pasting the results comparison before and after, or e2e results
• (Optional) The necessary documentation update, such as updating supported_models.md and examples for a new model.
• (Optional) Release notes update. If your change is user facing, please update the release notes draft in the Google Doc.

[chaunceyjiang]

How do you start vLLM?

This field structured_outputs_config.reasoning_parser can be set via the parameter --reasoning-parser qwen3.

[ahao-anyscale]

Is the user intended to supply these arguments? I seems that it would make more sense to automatically set them for reasoning models. I saw that gptoss had its reasoning-parser set automatically via verify_and_update_config. Thanks for reviewing!