CLIPTextEncodeLumina2_motorway_edition

Lumina2 is the open model line that leans into "prompting as instruction-following," and its text-encode node reflects that: there's a system prompt dropdown sitting right next to your actual prompt. This is the Motorway-ed clone of CLIPTextEncodeLumina2 - the CLIP comes out of the Motorway by key, you pick a system-prompt mode and type your text, and the conditioning goes back in under a key you name.

If you've never touched Lumina2, this node is your first look at how its conditioning is structured. The system_prompt field isn't a free-text box; it's a dropdown with two presets, and the author's own tooltip spells out the difference better than I can paraphrase:

• superior - "an assistant designed to generate superior images with the superior degree of image-text alignment"
• alignment - "an assistant designed to generate high-quality images with the highest degree of image-text alignment"

"Superior" and "alignment" read like near-synonyms, and they kind of are - the practical difference is a tuned tradeoff between image quality and how strictly the image matches your text. Treat the dropdown as a flavor switch: if your images aren't following the prompt tightly enough, flip to alignment; if they're over-literal and ugly, try superior. The two options exist because the model's own docs found neither one dominates.

How it works

The wrapper pulls one thing out of the Motorway and writes one back:

• INPUT_clip_key - key for the Lumina2 CLIP (default clip). Must be the right architecture for the model; a generic CLIP won't produce sensible conditioning.
• system_prompt - the dropdown above (superior / alignment).
• user_prompt - your actual prompt, multiline. The tooltip just says "The text to be encoded," which is refreshingly honest for this pack.
• OUTPUT_CONDITIONING_key - where the conditioning goes, default CONDITIONING.

The node instantiates the real core CLIPTextEncodeLumina2, encodes, and stores the result under the output key. The only visible output is the MOTORWAY 🚌💨; the sampler downstream reads the CONDITIONING key.

Motorway specifics

MOTORWAY 🚌💨 is required. The defaults assume your CLIP lives under clip; if your loader wrote it under CLIP, update the field or you'll hit the signature 'MotorwayClass' object has no attribute 'hash_' error, which just means you asked for a key nobody has written yet. Keys overwrite as they flow down the chain.

Install

cd ComfyUI/custom_nodes
git clone https://github.com/agilly1989/ComfyUI_agilly1989_motorway

Restart, or find "ComfyUI_agilly1989_motorway" in ComfyUI Manager. No pip dependencies; the Lumina2 model files are a separate download.

Fine print

This is an explicitly-beta pack - README opens with "BIG BROKEN WITH ASYNC WILL FIX WHEN I GET THE TIME/MOTIVATION" - and the current hotfix release has the _motorway_edition clone generation commented out of __init__.py, so on the newest main these nodes may not load at all. Workflows referencing them with missing nodes are hitting exactly that. GitHub issues are the channel; the author has said his Discord DMs get ignored.