How Blender Scripting Saves Hours: Automating 3D Asset Exports and Thumbnails

Blender scripting can seem intimidating at first, but once you start exploring it, the possibilities for saving time and improving results are endless. Recently, I’ve been diving into scripting while working on a 3D UI project for Decentraland’s In World Builder, and I’d like to share how it made my workflow smoother and more efficient. If you’re looking to automate tedious tasks and spend more time on the creative side of things, this one’s for you.

Here’s what Blender scripting helped us batch export— no manual exporting required 🎉

If you're completely new to Blender or Scripting, take a look at this post first:

Blender Scripting for Beginners: A Step by Step Guide.

---

Batch export individual assets with colliders using this Blender script.

The Situation: For this particular project, I was tasked with creating letters, numbers, and symbols for the In World Builder in Decentraland. Each character had different fonts and colorways, which resulted in a massive set of over 1600 files. As you can imagine, manually exporting each asset would’ve been a time-consuming nightmare.

Enter Blender Scripting: I knew there had to be a better way to handle the exporting process. So, I wrote a Blender script to export each asset individually, using the object’s name to keep things organized. On top of that, the script automatically added custom colliders for each asset to ensure they worked smoothly in Decentraland.

Once the exporting was done, I realized I needed thumbnails for each asset too. Instead of manually rendering 1600 images (no, thank you!), I wrote another script. This one created a thumbnail render for each individual asset and exported it as a PNG file with the same name.

Automating thumbnail creation for each asset with Blender scripting.

Not only did these scripts save me an incredible amount of time, but they also ensured consistency across all the assets.

Batch exporting and automatic thumbnail creation in action! Much quicker than doing it manually.

---

Learning as You Go: The best part about Blender scripting? You don’t have to be an expert to make it work for you. I’m far from a scripting guru, but by experimenting and problem-solving, I was able to drastically improve my workflow. The key is staying curious and being willing to learn—scripting is just another tool in your creative toolbox.

Ready to Try? If you’re interested in trying these scripts yourself, I’ve included them below. Feel free to tweak, adjust, and experiment! And if you have any questions, drop a comment or reach out—I’m always happy to chat about ways to work smarter in Blender.

Let’s keep learning together. What tasks have you automated in Blender, or what processes do you think could benefit from scripting? Share your ideas, and let’s see where this methodology can take us next.

Scripts:

• Script 1: Automating asset export with custom colliders.

batch_export_with_collider.py.zip

Download ZIP • 1KB

# Exports each selected object into its own file with an additional collider cube

import bpy

import os

# Export to blend file location

basedir = os.path.dirname(bpy.data.filepath)

if not basedir:

    raise Exception("Blend file is not saved")

# Define your custom set name here if needed

# set_name = "YourSetName"

view_layer = bpy.context.view_layer

obj_active = view_layer.objects.active

selection = bpy.context.selected_objects

bpy.ops.object.select_all(action='DESELECT')

for obj in selection:

    # Select the object

    obj.select_set(True)

    # Set the active object for some exporters

    view_layer.objects.active = obj

    # Get the object name and file path for export

    object_name = bpy.path.clean_name(obj.name)

    # If you want to add a set name prefix, uncomment the next line:

    # full_name = f"{set_name}_{object_name}"

    fn = os.path.join(basedir, object_name)

    # Create a simple cube for the collider

    bpy.ops.mesh.primitive_cube_add(size=0.25)

    collider = bpy.context.object

    collider.name = object_name + "_collider"

    collider.display_type = 'WIRE'

    # Position the collider at the object's origin

    collider.location = (obj.location.x, obj.location.y, obj.location.z + 0.15)

    # Parent the collider to the object

    collider.parent = obj

    # Select both the object and the collider for export

    collider.select_set(True)

    obj.select_set(True)

    # Export to GLB

    try:

        bpy.ops.export_scene.gltf(filepath=fn + ".glb", use_selection=True)

    except AttributeError:

        print("Error: glTF 2.0 Export Add-on is not enabled.")

        break

    # Deselect the objects

    collider.select_set(False)

    obj.select_set(False)

    # Delete the collider after export

    bpy.data.objects.remove(collider)

    print("Written:", fn + ".glb")

# Restore the previous active object

view_layer.objects.active = obj_active

# Reselect the original selection

for obj in selection:

    obj.select_set(True)

• Script 2: Thumbnail rendering for each asset in PNG format.

Create Thumbnail w Transparent BG.py.zip

Download ZIP • 2KB

import bpy

import os

# Define the resolution and file format for the thumbnails

thumbnail_size = 512

output_format = 'PNG'  # You can adjust this if needed

# Export to blend file location

basedir = os.path.dirname(bpy.data.filepath)

if not basedir:

    raise Exception("Blend file is not saved")

# Set render settings for thumbnails

bpy.context.scene.render.image_settings.file_format = output_format

bpy.context.scene.render.image_settings.color_mode = 'RGBA'  # RGBA for transparency

bpy.context.scene.render.resolution_x = thumbnail_size

bpy.context.scene.render.resolution_y = thumbnail_size

bpy.context.scene.render.film_transparent = True  # Make background transparent

bpy.context.scene.render.resolution_percentage = 100

# Add a sun lamp to the scene

def add_sunlight():

    bpy.ops.object.light_add(type='SUN', align='WORLD', location=(10, -10, 10))

    sun = bpy.context.object

    sun.data.energy = 5  # Adjust light intensity if needed

    sun.rotation_euler = (0.7854, 0, 0.7854)  # Rotate sun for better illumination

    return sun

# Set up the camera closer to the object

def setup_camera():

    bpy.ops.object.camera_add(location=(0, -0.36, 0.195))  # Closer camera

    camera = bpy.context.object

    camera.rotation_euler = (1.3, 0, 0)  # Adjust camera angle

    bpy.context.scene.camera = camera

    return camera

# Hide all objects except the current one

def hide_other_objects(exclude_object):

    for obj in bpy.context.scene.objects:

        if obj != exclude_object:

            obj.hide_render = True

# Restore visibility for all objects

def show_all_objects():

    for obj in bpy.context.scene.objects:

        obj.hide_render = False

# Get the view layer and selected objects

view_layer = bpy.context.view_layer

selection = bpy.context.selected_objects

for obj in selection:

    # Deselect all and select the current object

    bpy.ops.object.select_all(action='DESELECT')

    obj.select_set(True)

    view_layer.objects.active = obj

    # Hide all other objects from render

    hide_other_objects(obj)

    # Set up camera to frame the object

    camera = setup_camera()

    # Align the camera to fit the object (tightly frame the asset)

    bpy.ops.view3d.camera_to_view_selected()

    # Add sunlight for better illumination

    sun = add_sunlight()

    # Set the filename to the object name

    object_name = bpy.path.clean_name(obj.name)

    output_path = os.path.join(basedir, object_name + ".png")

    # Render the thumbnail and save the image

    bpy.context.scene.render.filepath = output_path

    bpy.ops.render.render(write_still=True)

    # Clean up by removing the camera and sun after render

    bpy.data.objects.remove(camera)

    bpy.data.objects.remove(sun)

    # Show all objects again

    show_all_objects()

    print("Thumbnail created:", output_path)

# Reselect the original selection

for obj in selection:

    obj.select_set(True)