ComfyUI Node: Text Generate Qwen3.5 (System Prompt)

Text Generate Qwen3.5 (System Prompt):

The TextGenerateQwen35SystemPrompt node is designed to facilitate advanced text generation tasks by leveraging the capabilities of the Qwen 3.5 model. This node is particularly useful for generating text based on a combination of user prompts and system messages, allowing for a structured and contextually rich output. It is ideal for scenarios where you need to inject a predefined system message before processing user input, thereby enhancing the coherence and relevance of the generated text. The node supports various input types, including text and images, and offers flexibility in controlling the length and sampling mode of the output. By integrating system prompts, it ensures that the generated text aligns with specific guidelines or themes, making it a powerful tool for creating detailed and context-aware content.

Text Generate Qwen3.5 (System Prompt) Input Parameters:

clip

This parameter represents the input clip, which is a required component for the node's operation. It serves as the foundational data upon which the text generation process is based.

prompt

The prompt parameter is a multiline string input where you can provide the user message. It supports all characters, including {} brackets, as the template uses concatenation instead of .format(). This allows for flexible and dynamic input without the risk of formatting errors.

system_message

The system_message is a multiline string input that allows you to inject a predefined message before the user input is processed. This message sets the context or guidelines for the text generation, and leaving it empty will skip the system block entirely.

image

This optional parameter allows you to input an image that can be used to influence the text generation process. If provided, the image can add context or visual cues that the model can incorporate into the generated text.

max_length

The max_length parameter controls the maximum number of tokens in the generated text. It accepts integer values ranging from 1 to 8192, with a default of 512. Adjusting this value allows you to manage the verbosity and detail level of the output.

sampling_mode

This parameter lets you select the sampling mode from a set of predefined options. The sampling mode influences how the model generates text, affecting factors like creativity and randomness in the output.

thinking

The thinking parameter is a boolean input that, when enabled, allows the model to operate in a "thinking" mode. If set to False, it suppresses the thinking process by wrapping it with ``\n</think>\n, which can be useful for controlling the flow and style of the generated text.

Text Generate Qwen3.5 (System Prompt) Output Parameters:

generated_text

The generated_text output parameter provides the final text generated by the node. This text is the result of processing the input parameters, including the prompt, system message, and any additional context provided by images or other inputs. It is designed to be coherent, contextually relevant, and aligned with the guidelines set by the system message.

Text Generate Qwen3.5 (System Prompt) Usage Tips:

• To achieve more contextually rich outputs, make sure to provide a detailed system_message that aligns with your desired theme or guidelines.
• Experiment with different sampling_mode options to find the balance between creativity and coherence that best suits your project needs.
• Use the max_length parameter to control the verbosity of the output, especially when generating longer narratives or detailed descriptions.

Text Generate Qwen3.5 (System Prompt) Common Errors and Solutions:

Missing clip input

• Explanation: The node requires a clip input to function correctly, and this error occurs when it is not provided.
• Solution: Ensure that you have included a valid clip input before executing the node.

Invalid max_length value

• Explanation: The max_length parameter must be within the specified range of 1 to 8192. An out-of-range value will trigger this error.
• Solution: Adjust the max_length parameter to a value within the acceptable range and try again.

Unsupported sampling_mode option

• Explanation: The sampling_mode parameter must be set to one of the predefined options. An invalid option will cause this error.
• Solution: Select a valid sampling_mode from the available options and re-run the node.