ComfyUI Node
Scene Describe (LM Studio)
Describe the photograph once and feed both the text proposer (`scene_hint`) and the caption builder. Without it the proposer invents signage for the wrong kind of place, or falls back on transcribing the nonsense already in the picture.
Scene Describe (LM Studio)
- image
- summary
- background
- setting
- report
◄base_urlhttp://localhost:1234/v1►
◄model_idqwen3-8b-vl-instruct-abliterated►
◄enabledtrue►
◄temperature0.00►
◄timeout180►
◄fallback_summary►
◄fallback_background►
◄fallback_setting►
CategoryFVMtools/signs
Inputs (9)
| Name | Type | Default | Description |
|---|---|---|---|
| image | IMAGE | The picture to read. | |
| base_url | STRING | http://localhost:1234/v1 | LM Studio, OpenAI-compatible. |
| model_id | STRING | qwen3-8b-vl-instruct-abliterated | Name it. 'Whatever is loaded' is not a default, it is a silent failure: LM Studio lists models it has not loaded, the call returns nothing, and the graph carries on with an empty description. Must be a VISION model. |
| enabled | BOOLEAN | true | Off: pass the fallbacks through unchanged, so the effect of the description can be measured. |
| temperature | FLOAT | 0.000–1 | 0.0. Measured over a hundred runs elsewhere in this project: at 0.2 and above the model invents one-off answers that never repeat. |
| timeout | INT | 18010–900 | — |
| fallback_summaryopt | STRING | — | |
| fallback_backgroundopt | STRING | — | |
| fallback_settingopt | STRING | — |
Outputs (4)
| Name | Type | Description |
|---|---|---|
| summary | STRING | — |
| background | STRING | — |
| setting | STRING | — |
| report | STRING | — |