ComfyUI Node
VLM Prompt Advanced (GGUF)
A ComfyUI node in VLMPrompt with 20 inputs and 1 output.
VLM Prompt Advanced (GGUF)
- image
- video
- prompt
◄model_nameGemma4-E4B-uncensored-Q4_K_M►
◄deviceauto►
◄preset_promptWan I2V Prompt►
◄system_prompt►
◄custom_prompt►
◄max_tokens768►
◄temperature0.60►
◄top_p0.90►
◄repetition_penalty1.20►
◄frame_count16►
◄ctx8192►
◄n_batch512►
◄gpu_layers-1►
◄image_max_tokens4096►
◄top_k0►
◄pool_size4194304►
◄keep_model_loadedfalse►
◄seed1►
CategoryVLMPrompt
Inputs (20)
| Name | Type | Default | Description |
|---|---|---|---|
| model_name | COMBO | Gemma4-E4B-uncensored-Q4_K_M | 4 options: Gemma4-E4B-uncensored-Q4_K_M, Qwen3-VL-8B-abliterated-Q4_K_M, Qwen3.5-4B-Uncensored-HauhauCS-Q4_K_M, Qwen3.5-9B-Uncensored-HauhauCS-Q4_K_M |
| device | COMBO | auto | 2 options: auto, cpu |
| preset_prompt | COMBO | Wan I2V Prompt | 27 options: NSFW I2V Tease, Wan I2V Prompt, Describe for Video, Annotate Image, Enhance Prompt, 🖼️ Tags, +21 |
| system_prompt | STRING | System prompt — sets the model's role/behavior. Leave empty for default. | |
| custom_prompt | STRING | User prompt — your instructions. Overrides preset if filled. | |
| max_tokens | INT | 76864–4096 | — |
| temperature | FLOAT | 0.600–2 | — |
| top_p | FLOAT | 0.900–1 | — |
| repetition_penalty | FLOAT | 1.200.5–2 | — |
| frame_count | INT | 161–64 | — |
| ctx | INT | 81921024–131072 | — |
| n_batch | INT | 51264–32768 | — |
| gpu_layers | INT | -1-1–200 | — |
| image_max_tokens | INT | 4096256–1024000 | — |
| top_k | INT | 00–32768 | — |
| pool_size | INT | 41943041048576–10485760 | — |
| keep_model_loaded | BOOLEAN | false | — |
| seed | INT | 11–4294967295 | — |
| imageopt | IMAGE | — | |
| videoopt | IMAGE | — |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| prompt | STRING | — |