LLM Output Parser

LLMOutputParser

ZImage/Utilities

Parse LLM output (JSON/YAML/text) to Z-Image encoder fields. Handles code fences, nested keys, and multi-turn conversations.

Pack: ComfyUI-QwenImageWanBridge

custom_nodes.ComfyUI-QwenImageWanBridge

Inputs (9)

Name	Type	Required
text_input	STRING	required
parse_mode	COMBO	required
user_prompt_key	STRING	optional
system_prompt_key	STRING	optional
thinking_key	STRING	optional
assistant_key	STRING	optional
fallback_to_passthrough	BOOLEAN	optional
strip_quotes	BOOLEAN	optional
previous_conversation	ZIMAGE_CONVERSATION	optional

Outputs (7)

Name	Type
user_prompt	STRING
system_prompt	STRING
thinking_content	STRING
assistant_content	STRING
raw_text	STRING
parse_status	STRING
conversation	ZIMAGE_CONVERSATION