How to Use Neural4D Studio: Create, Refine & Export

1. Get started in Studio

Open Neural4D Studio and sign in. Create an account if you are new. Free users receive 50 Credits each week.
Choose your starting point. On the Studio homepage, the AI 3D Agent lets you select Generate 3D, Generate Image, or Generate Video and describe what you need. You can add a reference image or mention a previous asset with @asset. You can also open a specific Studio tool directly.
Set visibility before you generate. Choose Public if you may want your work to appear in the Community for others to browse or Recreate. Private keeps the result visible only to you and requires a paid subscription.

2. Choose the right tool

Pick the input you have and the result you want. Image and video generation are separate from the 3D workflow. An image you generate can later become the input for Image to 3D.

Your starting point	Use	Result and key condition
A written description	Text to 3D	Generate a 3D model from a prompt.
One reference image	Image to 3D	Generate a 3D model from a photo or illustration. Batch mode accepts up to 10 images, producing a separate model for each.
A 3D idea you want to refine through conversation	Neural4D-2o	Generate a model inside 2o from text or an image, then refine it in that session. Models from the separate Text to 3D and Image to 3D tools cannot be transferred into 2o for editing. Each edit costs 30 Credits.
Six orthographic views	Multi-View to 3D	Reconstruct a model using six views of the same subject.
An anime character idea or reference image	AnimeArt	Create a customizable anime-style 3D character from text or an image, then export it as VRM.
A concept or reference image	Image Generation	Create a 2D image. A text prompt is required, even when you upload a reference image.
A video concept or reference image	Video Generation	Create a video. A text prompt is required; revise the prompt or reference image and generate again to change the result.
A part with exact dimensions	Coala	Create an industrial CAD model from text, a reference image, or a drawing. Current export is STL.

To explore more tools and see the latest feature descriptions, visit the Neural4D features directory.

For Image to 3D and Multi-View to 3D, supported uploads are PNG, JPG, JPEG, and WebP, up to 20 MB per image.

Neural4D Studio interface and tool selection

3. Create your first 3D model

Choose Text to 3D or Image to 3D. Write a clear prompt, or upload an image and describe the desired model. Use Multi-View to 3D when you have six orthographic views.
Set the geometry. Choose Triangle for detail or Quad for later editing and animation. Set a target polycount: 100,000 to 500,000 for Triangle, or 1,000 to 100,000 for Quad.
Set a pose for a humanoid character if needed. In Text to 3D and Image to 3D, Pose defaults to Auto. A-Pose and T-Pose are subscription features. Pose is not available in Batch Image to 3D or Multi-View to 3D.
Generate and inspect the result. Rotate, pan, and zoom the 3D viewer to check the shape. A base mesh takes about 90 seconds. PBR texturing is a separate step and takes additional time; a complete textured GLB takes more than two minutes in total.

For an anime-style character, open AnimeArt instead. Describe the character or add a reference image, customize its appearance and expression, then generate and export a VRM avatar. This is a separate workflow from the Text to 3D and Image to 3D settings above.

Before you submit: Credits are deducted when a task is submitted. A server-side failure returns the Credits automatically. Submitting the same request again creates another task and another charge.

Previewing generated models, images, and videos in Neural4D

4. Decide what to do with your 3D model

Add a material

AI Texture applies PBR-compatible textures to a model generated in 3D Studio. Use a text prompt or a 2D reference image. External mesh uploads are not supported. Each generation costs 20 Credits.

Animate a Studio Asset

In Animate, select an existing Studio Asset, rig it, apply a preset motion, and download an animated GLB. Models over 300K polygons need remeshing before rigging. Animate does not accept local model uploads or custom motions.

Create a multicolor model

Multicolor starts from a textured Neural4D model. It detects color regions and lets you adjust a palette of 1 to 8 colors. Generate for 30 Credits, then download the colored model as 3MF or GLB. The default download format is 3MF.

Split a model into parts

Model Segmentation starts from your completed Neural4D model, with or without texture. Smart Split creates a set of separate parts for printing. This subscription feature costs 40 Credits per generation; the default download is a 3MF containing all parts as untextured geometry.

Explore and Recreate

Browse the Community in Studio to preview shared models, images, and videos. Select Recreate to open the relevant workspace with the original prompt and settings prefilled, then make your own version.

Multicolor prepares color regions and Model Segmentation prepares separate parts. Neither performs slicing, support generation, G-code creation, or a printer task.

5. Find, manage, and export your assets

Open Assets Library to browse your Models, Images, and Videos. Search for an asset or mark it as a favorite. Batch Actions lets you select multiple assets to favorite, download as a ZIP, or delete. Deleted assets cannot be recovered.

Open an asset to review it and choose a supported download format. Formats depend on the tool that created the result:

Asset or workflow	Supported download format
Text to 3D, Image to 3D, AI Texture	OBJ, FBX, GLB, USDZ, STL, BLEND
Animate	GLB
AnimeArt	VRM
Multicolor	3MF, GLB
Model Segmentation	3MF by default, containing all current parts as untextured geometry
Coala	STL
Image Generation	PNG
Video Generation	MP4

Commercial use: Assets created under a paid subscription have commercial use rights. Free-plan models are marked Trial and are for testing only.

6. Check Credits and usage

Neural4D uses Credits for generation and some follow-up actions. Free users receive 50 Credits each week. For current Go, Plus, and Pro plans, see Pricing. Open Account to view Usage History, including the date, type, details, and amount for deductions or returns.

Usage History currently has no search, filter, or export control. A failed server task may appear as an incomplete task followed by a Credits Return entry.

Frequently asked questions

Can I upload a model to AI Texture or Animate?
AI Texture uses models generated in 3D Studio. Animate starts from a Studio Asset; it does not accept a local model upload in the Animate workspace.

Can Neural4D-2o edit a model from another tool?
No. Start a model inside Neural4D-2o using text or an image, then refine it in the same 2o session. Models from the separate Text to 3D and Image to 3D tools cannot be transferred into 2o for editing.

How do I change a generated video?
Edit the text prompt or replace the reference image, then generate again.

Where can I find my creations?
Open Assets Library and select Models, Images, or Videos. You can search, favorite, or batch download your assets.

How to use Neural4D Studio