[Usage]: How to turn off the deep thinking mode of the model

[m08594589-source]

System Info

System Information:

• OS:
• Python version:
• CUDA version:
• GPU model(s):
• Driver version:
• TensorRT-LLM version:

Detailed output:

Paste the output of the above commands here

How would you like to use TensorRT-LLM

How to add the cancellation of deep thinking in the command for starting the model?
trtllm-serve /app/models/Qwen3-32B-NVFP4 --host 0.0.0.0 --port 8000

Before submitting a new issue...

• Make sure you already searched for relevant issues, and checked the documentation and examples for answers to frequently asked questions.

[coderabbitai]

📝 CodeRabbit Plan Mode

Generate an implementation plan and prompts that you can use with your favorite coding agent.

• Create Plan

Examples

• Example 1
• Example 2

_{You can configure auto-planning by selecting labels in the issue_enrichment configuration.}

🔗 Similar Issues

Related Issues

• [Usage]: The multimodal feature in TensorRT-LLM is just a demo, and I see it only supports Qwen2-VL. If I want to deploy Qwen2.5-VL or Qwen3-VL, do I need to develop it myself? Will the official team not adapt it for new models going forward? #10069
• [Question] Can Qwen3-next be used with attention dp? #9397
• [Usage]: Which workflow will be support for tensorrt-LLM? The engine build path or the PyTorch backend/trtllm-serve? #7131

👤 Suggested Assignees

• ztzywm
• RobertLou
• JoJoLev

---

🧪 Issue enrichment is currently in open beta.

To disable automatic issue enrichment, add the following to your .coderabbit.yaml:

issue_enrichment:
  auto_enrich:
    enabled: false

💬 Have feedback or questions? Drop into our discord or schedule a call!