Passthrough node that instantly unloads any running ollama model mid workflow. No API calls, just accepts plain string as a model name and executes 'ollama stop ' command in your system.

Note that for the command to work the model name should be exactly as it appears in ollama lists:

• ❌️'qwen3-vl-abliterated'
• ❌️'qwen3-vl-abliterated:4b-instruct'
• ✅️'huihui_ai/qwen3-vl-abliterated:4b-instruct'