FAQ

1. Why doesn't the page show Lux3D 3.0 or G1?

The model list is delivered by server configuration. Confirm the current environment, account permissions, and page version; if it still doesn't appear after refreshing, provide your account and page information via in-app feedback.

2. Why can't G1 add multiple images?

You must be in "Image-to-3D" and select Lux3D-G1, and the server model configuration must enable multi-image capability. Switching into or out of G1 will clear the current images; please re-upload.

3. Why are images cleared after switching models?

G1 and other models have different input structures. The code actively clears images when crossing the G1 model boundary to prevent incompatible single-image/multi-image inputs from being carried into new tasks.

4. Why can't I start generation?

Check login, API key, credits, required inputs, upload status, model, size, concurrency limit, and image generation style. Material repaint also requires a target model, reference image, and available material pipeline.

5. Why does the task keep queuing?

During peak hours or when the concurrency limit is reached, tasks will queue. Do not repeatedly submit the same task; observe the status in the task panel and wait for running tasks to complete.

6. Why is the 3DGS preview empty?

Confirm that the task actually returned a G1 3DGS artifact, the artifact URL or asset ID is ready, and switch to 3DGS mode in the preview node. Regular GLB results should use the model preview or LuxReal.

7. Why is a certain export format unavailable?

Only formats actually returned by the task and in "ready" status will be displayed. Formats not requested, still being generated, or unsupported by the current model will not be available for download.

8. Why can't GLB/ZIP be previewed?

Local preview supports GLB, GLTF, or ZIP containing only one GLB. If it contains multiple GLBs, is missing the model file, or the structure doesn't meet requirements, unzip and check or use a supported file.

9. How can I get more stable results?

For single images, prefer images with a complete subject, clean background, and even lighting; for G1 multi-image, cover complementary angles of the same object; for transparent objects, avoid highlight overexposure; for text generation, specify the shape, material, color, and style.