Y7 LM Studio (Vision)

LM Studio (Vision)<div style="font-size: 0.8em"><div id="Y7_shortdesc" style="margin-bottom: 15px;">Send an image to a vision-capable LLM in LM Studio for analysis and description.</div></div><div style="font-size: 0.8em"><div style="margin-bottom: 8px; margin-left: 0px; font-size: 12px;">Connects to an LM Studio server and sends an image along with an instruction to a vision-language (VL) model. The model must be vision-enabled or an error will be raised.</div></div><div style="font-size: 0.8em"><div style="margin-bottom: 8px; margin-left: 0px; font-size: 12px;">Inputs:</div></div><div style="font-size: 0.8em"><div style="margin-bottom: 8px; margin-left: 20px; font-size: 12px;">- <code style="border: 1px solid #666; border-radius: 3px; padding: 0px 1px; font-family: monospace; display: inline-block;">image</code>: The image to analyze (required)</div></div><div style="font-size: 0.8em"><div style="margin-bottom: 8px; margin-left: 20px; font-size: 12px;">- <code style="border: 1px solid #666; border-radius: 3px; padding: 0px 1px; font-family: monospace; display: inline-block;">model_identifier</code>: The VL model name/identifier loaded in LM Studio (connect a Select LMS Model node or type manually)</div></div><div style="font-size: 0.8em"><div style="margin-bottom: 8px; margin-left: 20px; font-size: 12px;">- <code style="border: 1px solid #666; border-radius: 3px; padding: 0px 1px; font-family: monospace; display: inline-block;">system_message</code>: The instruction sent alongside the image (default describes the image in detail)</div></div><div style="font-size: 0.8em"><div style="margin-bottom: 8px; margin-left: 20px; font-size: 12px;">- <code style="border: 1px solid #666; border-radius: 3px; padding: 0px 1px; font-family: monospace; display: inline-block;">reasoning_tag</code>: Tag name used to extract reasoning blocks (e.g., <code style="border: 1px solid #666; border-radius: 3px; padding: 0px 1px; font-family: monospace; display: inline-block;">think</code> for <code style="border: 1px solid #666; border-radius: 3px; padding: 0px 1px; font-family: monospace; display: inline-block;"><think>...</think></code>)</div></div><div style="font-size: 0.8em"><div style="margin-bottom: 8px; margin-left: 20px; font-size: 12px;">- <code style="border: 1px solid #666; border-radius: 3px; padding: 0px 1px; font-family: monospace; display: inline-block;">ip</code> / <code style="border: 1px solid #666; border-radius: 3px; padding: 0px 1px; font-family: monospace; display: inline-block;">port</code>: LM Studio server address (default: localhost:1234)</div></div><div style="font-size: 0.8em"><div style="margin-bottom: 8px; margin-left: 20px; font-size: 12px;">- <code style="border: 1px solid #666; border-radius: 3px; padding: 0px 1px; font-family: monospace; display: inline-block;">temperature</code>: Controls randomness (0.01-1.0, default 0.7)</div></div><div style="font-size: 0.8em"><div style="margin-bottom: 8px; margin-left: 20px; font-size: 12px;">- <code style="border: 1px solid #666; border-radius: 3px; padding: 0px 1px; font-family: monospace; display: inline-block;">max_tokens</code>: Maximum tokens to generate (-1 for unlimited)</div></div><div style="font-size: 0.8em"><div style="margin-bottom: 8px; margin-left: 20px; font-size: 12px;">- <code style="border: 1px solid #666; border-radius: 3px; padding: 0px 1px; font-family: monospace; display: inline-block;">unload_llm</code>: Unload the LLM from LM Studio after generation</div></div><div style="font-size: 0.8em"><div style="margin-bottom: 8px; margin-left: 20px; font-size: 12px;">- <code style="border: 1px solid #666; border-radius: 3px; padding: 0px 1px; font-family: monospace; display: inline-block;">unload_comfy_models</code>: Free VRAM by unloading ComfyUI models before running the LLM</div></div><div style="font-size: 0.8em"><div style="margin-bottom: 8px; margin-left: 0px; font-size: 12px;">Outputs:</div></div><div style="font-size: 0.8em"><div style="margin-bottom: 8px; margin-left: 20px; font-size: 12px;">- <code style="border: 1px solid #666; border-radius: 3px; padding: 0px 1px; font-family: monospace; display: inline-block;">Response</code>: The model's analysis/description with reasoning blocks removed</div></div><div style="font-size: 0.8em"><div style="margin-bottom: 8px; margin-left: 20px; font-size: 12px;">- <code style="border: 1px solid #666; border-radius: 3px; padding: 0px 1px; font-family: monospace; display: inline-block;">Reasoning</code>: The extracted reasoning content (if present)</div></div><div style="font-size: 0.8em"><div style="margin-bottom: 8px; margin-left: 0px; font-size: 12px;">Requires the <code style="border: 1px solid #666; border-radius: 3px; padding: 0px 1px; font-family: monospace; display: inline-block;">lmstudio</code> Python package: <code style="border: 1px solid #666; border-radius: 3px; padding: 0px 1px; font-family: monospace; display: inline-block;">pip install lmstudio</code></div></div><div style="font-size: 0.8em"><div style="margin-bottom: 8px; margin-left: 0px; font-size: 12px;">Note: The loaded model must support vision. Non-vision models will raise an error.</div></div>