ComfyUI Node

Basic Ollama

Interface with Ollama LLMs, supporting dynamic image inputs.

By BobRandomNumber·Created 9 months ago·Updated 5 months ago· 1
Basic Ollama
  • images
  • text
  • thinking
prompt
ollama_modelLoading...
generation_seed0
enable_thinktrue
saved_sys_promptZ-Image
use_sys_prompt_belowfalse
system_prompt
CategoryOllama

Inputs (8)

NameTypeDefaultDescription
promptSTRINGThe main text prompt for the LLM.
ollama_modelCOMBOLoading...Select the Ollama model to use.
generation_seedINT00–18446744073709550000Seed for randomness. Change this to force a new generation.
enable_thinkBOOLEANtrueEnable reasoning/thinking for supported models.
saved_sys_promptCOMBOZ-ImageSelect a pre-defined system prompt.
use_sys_prompt_belowBOOLEANfalseIf True, use the custom system prompt below instead of the saved one.
system_promptSTRINGCustom system prompt to guide the LLM's behavior.
imagesoptCOMFY_AUTOGROW_V3Dynamic image inputs for multimodal models.

Outputs (2)

NameTypeDescription
textSTRINGThe text generated by the LLM.
thinkingSTRINGThe reasoning trace (if available).