Choose a generation method
Compare Text to 3D, Image to 3D, Multi-view, and 3D Agent to pick the generation method that fits your input, the precision you need, and how you like to work.
No single method is best for every project. The right choice depends on what you start from, how closely the model must match a reference, and how much creative guidance you want.
Compare the methods
The following table compares the four methods. The ratings are relative to each other.
| Factor | Text to 3D | Image to 3D | Multi-view | 3D Agent |
|---|---|---|---|---|
| Reference images | None | 1 image | 2 to 4 images | Optional |
| Precision | Medium | High | Highest | Medium |
| Speed | High | High | Medium | Medium |
| Creative freedom | Highest | Medium | Low | Highest |
| Consistency across generations | Low | High | Highest | High |
| Back-side accuracy | Medium | Medium | Highest | Medium |
| Batch production | Yes | Yes, with Batch Images to 3D | Limited, because each model needs its own set of views | Yes, in a conversation |
| Guided creative process | No | No | No | Yes |
| Answers to 3D creation questions | No | No | No | Yes |
When to choose each method
Choose Text to 3D when:
-
You're exploring and want to try several directions.
-
You have no reference image, only an idea.
-
You want to control the style and material through the prompt.
Choose Image to 3D when:
-
You have a concept image, product photo, or sketch.
-
You need the result to match a reference image closely.
-
You're turning product photos into 3D models for e-commerce.
Choose Multi-view when:
-
You have images of the same object from several sides.
-
You need an accurate model from every side.
-
You're digitizing a physical product.
Multi-view is an option in Image to 3D with Meshy 7.1. When you turn on Multi-view, you can add a left, back, and right view to the main image.
Choose 3D Agent when:
-
You want help deciding what to create.
-
You need a batch of concepts in one consistent style, for example a set of game props.
-
You know how to 3D print but not how to model.
-
You want to ask questions about 3D creation or printing as you work.
FAQ
- Should I use Text to 3D or Image to 3D?
- Use Text to 3D to explore concepts when you have no reference. Use Image to 3D when you have a photo or sketch and need the model to match it.
- What is Multi-view best for?
- Multi-view suits high-fidelity reconstruction when you have images of the same object from several sides, such as when you digitize a product. Turn it on in Image to 3D with Meshy 7.1, and add up to 3 views to the main image.
- When should I use 3D Agent instead?
- Use 3D Agent when you want help deciding what to create, need a batch of concepts in one style, or prefer a guided conversation.
- Which Meshy method is most accurate?
- Image to 3D and Multi-view give the most precise results because they work from real visual references.
- Is there one best 3D generation method?
- No. The right choice depends on your input, how precise the model must be, and how much creative guidance you want.
Next steps
-
To generate a model from a written description, see Text to 3D.
-
To generate a model from one or more images, see Image to 3D.
-
To create a model through a conversation, see 3D Agent.