ComfyUI Node
Qwen 3.5 (WaveSpeed API)
A ComfyUI node in Qwen3.5 with 11 inputs and 2 outputs.
Qwen 3.5 (WaveSpeed API)
- image
- RESPONSE
- THINKING
◄modelQwen3.5-27B►
◄promptDescribe this image in detail.►
◄system_prompt►
◄max_tokens4096►
◄temperature1.00►
◄top_p0.95►
◄top_k20►
◄thinkingtrue►
◄api_key►
◄image_url►
CategoryQwen3.5
Inputs (11)
| Name | Type | Default | Description |
|---|---|---|---|
| model | COMBO | Qwen3.5-27B | WaveSpeed model. 35B-A3B: $0.16/$1.30 per M tokens. 27B: $0.20/$1.60. 122B-A10B: $0.26/$2.10. 397B-A17B: $0.39/$2.30 (best quality). |
| prompt | STRING | Describe this image in detail. | Text prompt for the model |
| system_prompt | STRING | Optional system prompt | |
| max_tokens | INT | 409664–81920 | Maximum tokens to generate |
| temperature | FLOAT | 1.000–2 | Sampling temperature. Thinking: 1.0, instruct: 0.7 |
| top_p | FLOAT | 0.950–1 | Nucleus sampling. Thinking: 0.95, instruct: 0.8 |
| top_k | INT | 201–100 | Top-K sampling |
| thinking | BOOLEAN | true | Enable thinking/reasoning mode (reasoning_effort: medium). Disable for faster, cheaper responses. |
| api_key | STRING | WaveSpeed API key. Leave empty to use WAVESPEED_API_KEY env var. | |
| imageopt | IMAGE | Single image input (resized + base64 encoded) | |
| image_urlopt | STRING | Image URL — sent directly to API, no base64. Preferred over image input. |
Outputs (2)
| Name | Type | Description |
|---|---|---|
| RESPONSE | STRING | — |
| THINKING | STRING | — |