How to Create 3D Models in Seconds with Meshy AI v6 (Step‑by‑Step Guide)

AI is changing how 3D assets are created. Instead of spending hours modeling from scratch, you can now type a short description or upload an image and get a ready-to-use 3D model in minutes. Meshy AI v6 is one of the leading tools in this space, and you can start using it for free.

Below is a full walkthrough of how to go from prompt or image to a textured, optimized 3D model you can render or even 3D print.

Getting Started with Meshy AI v6

Meshy AI is an AI-powered 3D modeling platform that can generate models from text or images. It’s useful for architectural visualization, game assets, 3D printing, and any project that needs quick, decent-quality 3D objects.

To get started:

1. Go to the Meshy AI website.

2. Create a free account (takes about a minute).

3. You’ll receive 200 free credits per month, which is enough to test and build several models.

Once you’re in, head to the Workspace. You’ll see two main options:

• Text-to-3D – generate a model from a written description.

• Image-to-3D – generate a model from a reference image.

How to Create a 3D Model from an Image

Image-to-3D is ideal when you already know what you want and have a clear visual reference. Meshy can turn that 2D reference into a 3D object with textures and geometry.

1. Choose the AI Model

In the image-to-3D workflow, you’ll first pick an AI model:

• Nano Banana – fastest, but least accurate.

• Nano Banana 2 – still fast, but more accurate and generally a better balance.

Select Nano Banana 2 for a good mix of speed and quality.

2. Write a Clear Prompt

Even with an image workflow, you’ll still describe what you want. For example:

“Fiddle leaf fig tree in a round minimalist concrete pot, white background.”

The mention of a white background is important. It tells the AI to focus only on the object and not generate an environment around it, which makes the model cleaner and easier to reuse in different scenes.

Set the number of generations to 4 and use a 1:1 aspect ratio to get multiple variations to choose from.

3. Send the Image to 3D

After Meshy generates the image variations, pick the one you like best and click “Send image to 3D”. A new window opens with generation settings.

You can usually leave the defaults as they are and hit Generate. Meshy will create a clay-like 3D model first.

Use your mouse to rotate, pan, and zoom to inspect the model. There’s also a geometry view so you can see the underlying mesh and polygon density.

4. Add Textures and PBR Maps

Next, switch to the Texture tab. Meshy will use your selected image as a reference for the texture.

Make sure these options are enabled:

• Remove lighting – strips baked-in lighting from the texture so it works better in different scenes.

• Generate PBR maps – creates physically based rendering maps (like normal, roughness, and metallic), which makes the material behave realistically in modern render engines.

Once textures are generated, you’ll see them listed in the top-right corner. The model should now look much closer to a final asset.

5. Remesh for Better Topology

By default, the generated model can be quite dense and messy in terms of geometry. To make it more usable in 3D software or games, use the Remesh feature.

Key settings:

• Polygon count – you can let Meshy decide with Adaptive, or choose Fixed and set a target like 30K polygons for a lighter model.

• Topology – choose Quad for cleaner, quad-based geometry that’s easier to edit and deform, even if it sacrifices a bit of detail.

Click Confirm to process the optimized version. After a minute or so, you’ll get a new, lighter model with better topology.

6. Export and Render

When you’re happy with the result, click Download at the bottom of the page and choose your preferred format, such as FBX, OBJ, or others depending on your pipeline.

Import the file into your 3D software or renderer. In the example workflow, the model is imported into Lumion for architectural rendering. Once placed in a scene, you can adjust lighting, camera angles, and materials as needed, then render a final image or animation.

If you’re interested in building full animated scenes with AI-generated assets, you may also want to check out this guide on creating 3D AI-animated videos with a single prompt.

How to Create a 3D Model from Text

Text-to-3D is even more direct: you describe what you want, and Meshy builds it from scratch. This is great for quickly filling scenes with props, environment pieces, or stylized objects.

1. Set Up a Text Prompt

In the Workspace, switch to Text-to-3D. Type a clear description of your object, for example:

“Stone garden fountain.”

Keep prompts specific enough to guide style and shape, but not so long that they become confusing. You can always regenerate with variations.

2. Choose the AI Model and Options

For text-to-3D, select the Standard Mesh 6 model, which is the latest and most capable version for general use.

Other options:

• Pose – this is only relevant for characters. For props and objects like fountains, set it to None.

• Generations – choose 4 to get multiple design variations.

Click Generate and wait for Meshy to produce the options.

3. Texture the Generated Model

Once you see the four generated versions, pick your favorite and head to the Texture tab. Click the button to generate textures.

After a short wait, you’ll have a fully textured 3D model. Again, you can inspect it from all angles and check the maps. From here, you can remesh and export just like in the image-to-3D workflow.

Import the model into your 3D software (such as Lumion, Blender, Unreal, or Unity) and render it inside your scene.

From AI Model to 3D Print

Meshy AI isn’t just for digital renders. You can also take models into a 3D printing workflow.

In the community section, you’ll find ready-made characters and objects shared by other users. Many of these have direct export options to popular 3D printers.

If you have your own printer, the general process is:

1. Download the model as an STL file.

2. Open it in your slicer or printer software.

3. Set print settings (layer height, infill, supports, etc.).

4. Send the file to your 3D printer.

Depending on size and quality settings, prints can take a few hours. The result is a physical object generated from an AI-designed 3D model, which is especially useful for prototypes, figurines, and custom parts.

Pricing, Credits, and Practical Use

Meshy AI offers a generous free tier:

• 200 credits per month to generate new models.

• Optional Pro plan starting at around $1 for the first month (limited-time offer), which unlocks more generations and features.

While not every generation is perfect, the tool is already strong enough to transform workflows, especially when you need lots of smaller assets to populate a scene—plants, props, background objects, and quick concept pieces.

If you’re building more complex AI workflows around 3D or automation, you might also find it useful to explore how to set up your first AI agent to manage repetitive tasks around asset generation and organization.

Used well, Meshy AI v6 can save hours of manual modeling, letting you focus more on design, composition, and storytelling rather than starting every object from a blank mesh.