Minimal entry node for persistent local GGUF chat sessions (config-driven defaults). — ComfyUI node from ComfyUI-LLM-Session. Run it in the cloud, no install…

Local LLM session nodes for ComfyUI using GGUF and llama.cpp, supporting Llama, Mistral, Qwen, DeepSeek, GLM, Gemma, Phi, LLaVA and gpt-oss, enabling both user–model chat and model-to-model dialogue without external runtimes like Ollama.

ComfyUI-LLM-Session

user_text

session_id

model

mmproj

history_dir

media

config_path

Minimal entry node for persistent local GGUF chat sessions (config-driven defaults).

Name	Type	Default	Description
user_text	STRING		User message for this turn
session_id	STRING	default	Session ID (maps to a history file). Same ID continues the chat.
model	COMBO	(No GGUF models found in models/LLM/)	GGUF model file in models/LLM/
mmproj	COMBO	(Auto-detect)	Manual selection is recommended.
history_dir	STRING		Optional directory for history/caches. Empty uses output/llm_session_sessions/.
mediaopt	*		Optional IMAGE tensor/batch or AUDIO input for this turn only (never saved to history)
config_pathopt	STRING		Optional override path to simple_defaults.json (advanced).

LLM Session Chat (Simple)

Inputs (7)

Outputs (1)