ComfyUI Node
Ollama Prompts Generator Tlant
A ComfyUI node in prompts with 6 inputs and 1 output.
Ollama Prompts Generator Tlant
- STRING
◄positive_text—►
◄reference_text—►
◄negative_text—►
◄prompt_templateAs a professional art critic, describe the image based on the descriptions provided in the following three reference paragraphs.
The first reference paragraph:
{positive_text}
The second reference paragraph:
{reference_text}
The third reference paragraph:
{negative_text}
Create a coherent and realistic scene in a single paragraph. Include:
Describe placement
Main subject details
Artistic style and theme
Setting and narrative contribution
Lighting characteristics
Color palette and emotional tone
Camera angle and focus
The elements described in the first reference paragraph must be included in the output paragraph, and if there is a conflict between the first reference paragraph and the second reference paragraph, the first reference paragraph takes precedence.
The description in the third reference paragraph contains negative elements and should not appear in the output paragraph. If no description for the third reference paragraph is provided, it can be ignored.
Merge image concepts if there is more than one.
Always blend the concepts, never talk about splits or parallel.
Do not split or divide scenes, or talk about them differently - merge everything to one scene and one scene only.
Blend all elements into unified reality. Use image generation prompt language. No preamble, questions, or commentary.►
◄custom_modelllama3.1:8b►
◄ollama_urlhttp://localhost:11434/api/generate►
Categoryprompts
Inputs (6)
| Name | Type | Default | Description |
|---|---|---|---|
| positive_text | STRING | — | |
| reference_text | STRING | — | |
| negative_text | STRING | — | |
| prompt_template | STRING | As a professional art critic, describe the image based on the descriptions provided in the following three reference paragraphs. The first reference paragraph: {positive_text} The second reference paragraph: {reference_text} The third reference paragraph: {negative_text} Create a coherent and realistic scene in a single paragraph. Include: Describe placement Main subject details Artistic style and theme Setting and narrative contribution Lighting characteristics Color palette and emotional tone Camera angle and focus The elements described in the first reference paragraph must be included in the output paragraph, and if there is a conflict between the first reference paragraph and the second reference paragraph, the first reference paragraph takes precedence. The description in the third reference paragraph contains negative elements and should not appear in the output paragraph. If no description for the third reference paragraph is provided, it can be ignored. Merge image concepts if there is more than one. Always blend the concepts, never talk about splits or parallel. Do not split or divide scenes, or talk about them differently - merge everything to one scene and one scene only. Blend all elements into unified reality. Use image generation prompt language. No preamble, questions, or commentary. | — |
| custom_modelopt | STRING | llama3.1:8b | — |
| ollama_urlopt | STRING | http://localhost:11434/api/generate | — |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| STRING | STRING | — |