Turn an image into a 3D model

01　Choose an image with a clear subject

Use an image with a complete subject, clear edges, and as little obstruction as possible. AI infers the parts it cannot see, so you need to inspect the result.

02　Upload the image and describe your goal

Add a reference image using the plus button. Wait for the upload, then enter a request such as “Turn this image into 3D and keep its shape.” The reference cards show the images for this request.

You can describe the task without choosing a capability first. Or select Image to 3D to adjust available settings. Generation access during beta depends on your account.

03　Review the proposal and credits

Review conversation cost information before sending. When the assistant proposes a generation task, check the images, settings, and generation credits. A generation task starts only after you confirm its proposal.

Follow progress in the conversation. You can leave or refresh the page and return through conversation history without submitting the same task again.

04　Preview, download, or create another version

Once the result is ready, open the preview and rotate or zoom the model. Completed models are also available in My works.

The archived original is a GLB file. If STL export is enabled for that model, review its separate quote before converting. Creating another version requires a new confirmation. Check dimensions and structure before manufacturing or printing.