> ## Documentation Index
> Fetch the complete documentation index at: https://mcp-for-blender.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Scientific Visualization and Data Scenes with Blender MCP

> Build accurate 3D scientific visualizations, molecular models, geographic representations, and data-driven scenes in Blender using AI-generated Python code.

Blender MCP is well suited for scientific visualization because it can execute arbitrary Python code in Blender. Describe what you want to visualize and Claude writes and runs the `bpy` code to build it. This makes it possible to create custom geometry, data-driven shapes, and precise models that go beyond pre-built asset libraries.

## Key Workflows for Scientific Visualization

### Custom Geometry with `execute_blender_code`

The `execute_blender_code` tool is the core of scientific workflows. Claude writes Blender Python to generate:

* Molecular structures (DNA double helix, protein backbones, crystal lattices)
* Terrain meshes from height maps or mathematical functions
* Data-driven shapes (3D scatter plots, bar charts, vector fields)
* Cross-section and layered models (Earth's layers, anatomical slices)

### Realistic Materials and Lighting from Poly Haven

Once geometry is built, apply Poly Haven PBR materials and HDRIs to make the visualization publication-ready. Stone, metal, glass, and organic textures are all available.

### Iterative Refinement with Scene Inspection

Use `get_scene_info` to inspect object names, locations, and properties. Use `get_viewport_screenshot` to see the current state and guide Claude toward corrections.

## Example Prompts

| Prompt                                               | What Claude builds                                                     |
| ---------------------------------------------------- | ---------------------------------------------------------------------- |
| "Create a 3D model of a DNA double helix"            | Custom `bpy` geometry with two intertwined strands and base pairs      |
| "Generate a terrain mesh from a height map"          | Python script that reads image data and displaces a plane mesh         |
| "Visualize a 3D scatter plot from this data"         | Script that places spheres or cubes at data coordinates                |
| "Create a cross-section model of the Earth's layers" | Concentric spheres with distinct materials for core, mantle, and crust |
| "Build a protein backbone from these coordinates"    | Cylinders and spheres connected by script                              |

<Note>
  The code execution capability means you can ask Claude to write custom Blender Python for any visualization need, not just pre-built tools. If you can describe it, Claude can script it.
</Note>

## Workflow Tips

<Tip>
  Break complex visualizations into steps: build geometry first, then materials, then lighting. After each step, use `get_viewport_screenshot` to verify progress before moving on.
</Tip>

<Tip>
  Save your Blender file before running complex `execute_blender_code` prompts. This gives you a safe point to return to if the generated code needs adjustment.
</Tip>

## Recommended Tools

<CardGroup cols={3}>
  <Card title="execute_blender_code" icon="code">
    The most important tool for scientific workflows. Generates custom geometry, data imports, and precise object manipulation.
  </Card>

  <Card title="get_scene_info" icon="info">
    Inspect object names, coordinates, and properties to verify accuracy.
  </Card>

  <Card title="get_viewport_screenshot" icon="image">
    Visual feedback for iterative refinement of complex models.
  </Card>
</CardGroup>

## Next Steps

* [Robotics simulation environments](/use-cases/robotics-simulation)
* [Architecture visualization](/use-cases/architecture)
* [Complete scene generation](/use-cases/scene-generation)