ComfyUI Node: Qwen 3.5 (WaveSpeed API)

Qwen 3.5 (WaveSpeed API):

The Qwen35WaveSpeed node is designed to interface with the Qwen3.5 model via the WaveSpeed API, allowing you to leverage powerful language model capabilities without the need for a local GPU. This node is particularly beneficial for tasks such as captioning, where it utilizes WaveSpeed's infrastructure to deliver efficient and effective results. By using the WaveSpeed API, you can access a range of Qwen3.5 models, each tailored for different performance and cost needs, ensuring flexibility and scalability in your AI projects. The node supports advanced sampling techniques and reasoning modes, making it a versatile tool for generating high-quality language outputs.

Qwen 3.5 (WaveSpeed API) Input Parameters:

top_p

The top_p parameter controls nucleus sampling, which is a technique used to limit the number of potential next words by considering only the most probable ones. This parameter ranges from 0.0 to 1.0, with a default value of 0.95. A higher top_p value allows for more diverse outputs, while a lower value results in more focused and deterministic responses. For general thinking tasks, a value of 0.95 is recommended, whereas for instructional tasks, 0.8 might be more suitable.

top_k

The top_k parameter specifies the number of top probable words to consider during the sampling process. It is an integer value ranging from 1 to 100, with a default of 20. By adjusting top_k, you can control the randomness of the output; a lower value results in more predictable outputs, while a higher value increases variability and creativity in the responses.

thinking

The thinking parameter is a boolean that enables or disables the reasoning mode of the model. When set to True, the model engages in a medium level of reasoning effort, which can enhance the quality and depth of the responses. Disabling this option can lead to faster and more cost-effective outputs, suitable for scenarios where speed is prioritized over depth.

api_key

The api_key parameter is a string that holds your WaveSpeed API key, which is necessary for authenticating requests to the WaveSpeed API. If left empty, the node will attempt to use the WAVESPEED_API_KEY environment variable. This parameter is crucial for accessing the API and ensuring secure communication with the WaveSpeed infrastructure.

Qwen 3.5 (WaveSpeed API) Output Parameters:

output_text

The output_text parameter provides the generated text from the Qwen3.5 model. This output is the result of processing the input parameters and utilizing the selected model via the WaveSpeed API. The text can vary in style and content based on the input settings, offering a wide range of possibilities for creative and informative outputs.

Qwen 3.5 (WaveSpeed API) Usage Tips:

• To achieve more creative and diverse outputs, consider increasing the top_p and top_k values. This will allow the model to explore a wider range of possibilities in its responses.
• If you require faster responses and are less concerned with the depth of reasoning, disable the thinking mode to reduce processing time and cost.

Qwen 3.5 (WaveSpeed API) Common Errors and Solutions:

Missing API Key

• Explanation: This error occurs when the API key is not provided or the WAVESPEED_API_KEY environment variable is not set.
• Solution: Ensure that you have entered a valid API key in the api_key parameter or set the WAVESPEED_API_KEY environment variable before running the node.

Invalid Parameter Value

• Explanation: This error arises when one or more input parameters are set outside their allowed range or with incorrect data types.
• Solution: Double-check the values of top_p, top_k, and other parameters to ensure they fall within the specified ranges and are of the correct type. Adjust them accordingly and try again.