Nodes/WAS LMStudio Easy-Query/LM Studio Options
ComfyUI Node

LM Studio Options

A ComfyUI node in LM Studio with 10 inputs and 1 output.

By WASasquatch·Created 9 months ago·Updated 5 months ago· 17
LM Studio Options
    • options
    temperature0.20
    max_tokens512
    seed0
    top_p1.00
    top_k0
    frequency_penalty0.00
    presence_penalty0.00
    repeat_penalty1.00
    strip_thinking_tagsfalse
    stop
    CategoryLM Studio

    Inputs (10)

    NameTypeDefaultDescription
    temperatureFLOAT0.200–2Sampling temperature. Higher = more random; lower = more deterministic.
    max_tokensINT512-1–32768Maximum number of new tokens to generate for the response. Use -1 for unlimited (recommended for thinking models).
    seedINT00–2147483647Optional random seed for reproducible sampling. Use 0 to disable and let the model choose.
    top_pFLOAT1.000–1Nucleus sampling: consider tokens with cumulative probability up to top_p. Set to 1.0 to disable.
    top_kINT00–2048Top-K sampling: only consider the top_k most likely tokens. Set to 0 to disable.
    frequency_penaltyFLOAT0.000–2Penalize tokens proportionally to how often they have appeared. Helps reduce repetition.
    presence_penaltyFLOAT0.000–2Penalize tokens if they have appeared at all. Encourages introducing new topics.
    repeat_penaltyFLOAT1.000–2Generic repetition penalty (model/engine specific). 1.0 means no penalty.
    strip_thinking_tagsBOOLEANfalseRemove <thinking>...</thinking> tags and their content from the output. Useful for getting clean captions without reasoning text.
    stopoptSTRINGComma-separated stop strings. Generation will stop when any is encountered. Leave blank for none.

    Outputs (1)

    NameTypeDescription
    optionsLMSTUDIO_OPTIONS