Z-Image Turbo + Ollama - Simple Workflow

I made this workflow because Z-Image Turbo works best with detailed prompts with 500+ words.

This simple Workflow will enhance your Prompt using your favorite Ollama Model. Useful also if you use a multilingual Ollama model to translate from any language.

Version 2.0

So this is a major revamp of my previous versions.

This now includes:

• Ollama Enhance Prompt

• Use a Rereference image for either or both together of

  • Ollama Reference Image (+ Prompt)

  • Controlnet

• Change a Person using a Character Lora, so you can have 2 different Characters in one image, what is otherwise not possible with ZIT (inspired by https://civitai.com/models/2232773/2-lora-characters-z-image)

• SeedVR2 Upscaler

There are notes in the workflow to explain most of the important things.

Note:

I use scheduler beta57, which is not by default available in "Detailer (SEGS)" note, but you can easily add it, simply add it to the "ADDITIONAL_SCHEDULERS" list in the file custom_nodes/comfyui-impact-pack/core.py, so the line read as:

ADDITIONAL_SCHEDULERS = ['AYS SDXL', 'AYS SD1', 'AYS SVD', 'GITS[coeff=1.2]', 'LTXV[default]', 'OSS FLUX', 'OSS Wan', 'OSS Chroma', 'beta57']

Version 1.1

Add option to use a reference image for Ollama (make sure to use a Vision Model).

The first example image uses the second example image as reference, adding and changing details using Ollama.

Version 1.0

This is the base version.

The first example image is made with Ollama Enhance, second is the same prompt without Ollama.