comfyui-local-ollama-translator

Local Ollama translation node for ComfyUI with English/Simplified Chinese prompt translation and privacy-first localhost defaults.

By InfernusIntraMe·Created 2 months ago·Updated 3 days ago· 2

InfernusIntraMe/ComfyUI-Local-Ollama-Translator

Nodes—

On cloudLocal install

Stars2

Updated3 days ago

Readme

ComfyUI Local Ollama Translator

A local-first ComfyUI custom node for English ↔ Simplified Chinese translation using a locally running Ollama model.

It is designed for prompt translation inside ComfyUI workflows. The node sends text to Ollama on your own machine, receives the translation, shows a GUI preview, and outputs the translated text as a STRING that can be wired into CLIP Text Encode or other text-input nodes.

Features

English → Simplified Chinese
Chinese → American English
Auto direction mode
Image Prompt mode for prompt-style text
Natural, Literal, and UI Text modes
Preserve-terms field for model names, LoRA triggers, brands, file paths, or technical tokens
GUI translation preview
translated_text and source_text outputs
Privacy-first default endpoint: http://127.0.0.1:11434
Remote Ollama endpoints blocked by default
No extra Python dependencies

Privacy

By default, this node only talks to Ollama running on the same machine:

http://127.0.0.1:11434

It does not call Google Translate, DeepL, ChatGPT, or any external translation API.

The allow_remote_endpoint option is off by default. Leave it off unless you intentionally want to send text to another machine running Ollama.

Requirements

ComfyUI
Ollama installed and running
A multilingual Ollama model

Recommended model:

ollama pull qwen2.5:7b

Other multilingual models can work too. Set the node's model field to match your installed Ollama model name.

Installation

Clone this repository into your ComfyUI custom nodes folder:

cd /d path\to\ComfyUI\custom_nodes
git clone https://github.com/InfernusIntraMe/ComfyUI-Local-Ollama-Translator.git

Example:

cd /d D:\AI\ComfyUI\custom_nodes
git clone https://github.com/InfernusIntraMe/ComfyUI-Local-Ollama-Translator.git

Restart ComfyUI.

The node appears under:

Local/Ollama/Local Ollama Translator

Basic usage

For translation only, place the node on the canvas, enter text, and queue the workflow. The translation appears in the node preview and the translated_text output.

For image generation, wire it like this:

Local Ollama Translator: translated_text
        ↓
CLIP Text Encode: text

Your actual image workflow still needs its own proper text encoder:

Your workflow's Load CLIP / text encoder
        ↓
CLIP Text Encode: clip

This node translates prompt text. It does not replace the model's text encoder.

Recommended image-prompt settings

direction: English → Simplified Chinese
style: Image Prompt
model: qwen2.5:7b
endpoint: http://127.0.0.1:11434
keep_alive: 0
allow_remote_endpoint: false
debug_logging: false

`num_ctx`

num_ctx is the Ollama context length, not a seed. It controls how much text the model can see in one run.

For short prompt translation, 8192 is usually enough. The default 32768 gives more room for long prompts, captions, lyrics, or larger text blocks.

Notes for workflow sharing

ComfyUI workflows can store prompt text, translations, filenames, and widget values inside the JSON. Clear private prompts and generated preview text before uploading workflows publicly.

License

MIT