UNI-1 Max
The bigger-budget sibling for when the first pass isn't enough
- seedImage
- referenceImages
- advancedFeatures.watermark.image
- image
UNI-1 Max is the answer to the question everyone asks after a good-but-not-great UNI-1 render: "can it do better?" It's the higher tier of Luma's image model, same interface, same web-search trick, more of whatever the base model was holding back. In the hosted-model world "more" usually means resolution headroom and detail budget, and "Max" is where you spend it when the composition finally works.
Like its base sibling, it's a closed Luma model rented per generation through Runware's cloud in the ComfyUI-Runware pack. The local community's verdict on closed hosted image models is short and unflattering - "not local, not open source" - but if you're already in this pack, that ship has sailed. The question is which tier to queue, and Max is the tier for the keeper shot, not the first draft.
How it works
Same mechanism as every node here: model AIR luma:uni@1-max, imageInference task, REST request through the runware-sdk, native IMAGE output. Inputs mirror UNI-1 exactly:
positivePrompt- required.seedImage/referenceImages- IMAGE conditioning sockets.settings.search- web search for visual references before generating; use it for real-world subjects, landmarks, products.settings.style- preset, defaultauto.width/height- 1024;numberResults1–4;watermarksocket optional.
The output
One image socket, type IMAGE, into Preview Image or Save Image.
Installing it
The pack install covers every node, once:
cd ComfyUI/custom_nodes
git clone https://github.com/Runware/ComfyUI-Runware
pip install -r ComfyUI-Runware/requirements.txt
Restart, set your key via ComfyUI Settings → Runware API key, RUNWARE_API_KEY, or runware auth login.
Where people get burned
The tier's entire value proposition is "pay more for more," so the honest advice is: don't start here. Run the idea through Runware_luma_uni_1 to lock the composition, then re-queue the winner on Max - otherwise you're paying Max prices for prompt roulette. And remember the meter is the real user interface for hosted models: the node prints cost per run, and numberResults is the quiet way to triple that bill. If the extra spend isn't visibly buying detail, drop back to base and save the credits.
Inputs (21)
| Name | Type | Default | Description |
|---|---|---|---|
| positivePrompt | STRING | Text prompt describing elements to include in the generated output. | |
| seedImageopt | IMAGE | — | |
| referenceImagesopt | IMAGE | — | |
| advancedFeatures.watermark.imageopt | IMAGE | — | |
| sizeopt | COMBO | 3K (1:1) | Output size. Only one of resolution or dimensions is sent. |
| numberResultsopt | INT | 11–20 | Number of results to generate. Each result uses a different seed, producing variations of the same parameters. |
| advancedFeaturesopt | BOOLEAN | false | Enable to set advancedFeatures. Off uses the model's default. |
| advancedFeatures.watermark.bgColoropt | STRING | Background color in hex format. | |
| advancedFeatures.watermark.displayPositionopt | COMBO | (default) | Watermark position. |
| advancedFeatures.watermark.fontColoropt | STRING | Text color in hex format. | |
| advancedFeatures.watermark.opacityopt | BOOLEAN | false | Enable to set advancedFeatures.watermark.opacity. Off uses the model's default. |
| advancedFeatures.watermark.opacity_valueopt | FLOAT | 0.100.1–1 | Watermark opacity. |
| advancedFeatures.watermark.textopt | STRING | Watermark text. | |
| safetyopt | BOOLEAN | false | Enable to set safety. Off uses the model's default. |
| safety.checkContentopt | BOOLEAN | false | Enable or disable content safety checking. |
| settings.searchopt | BOOLEAN | false | When enabled, the model searches the web for visual references before generating. This improves accuracy for prompts that reference real-world subjects, landmarks, or products. |
| settings.styleopt | COMBO | auto | Applies a style preset to the generation. |
| ttlopt | BOOLEAN | false | Enable to set ttl. Off uses the model's default. |
| ttl_valueopt | INT | 60 | Time-to-live (TTL) in seconds for generated content. Only applies when `outputType` is `URL`. |
| outputFormatopt | COMBO | JPG | File format for the generated image. |
| outputQualityopt | INT | 9520–99 | Compression quality of the output. Higher values preserve quality but increase file size. |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| image | IMAGE | — |