Description
Common Sense Machines (CSM) is an AI platform that generates 3D models and worlds from images, text, or sketches. It uses deep learning to infer geometry, texture, and materials from minimal inputs. The platform offers a credit-based system with quality tiers — generation costs 8-40 credits depending on AI model and quality. In January 2026, CSM was acquired by Google DeepMind, signaling deeper integration with Google's spatial AI offerings. Before the acquisition, plans included Tinkerer (free, ~10 generations), Starter ($50/year), Maker ($200/year), and Creative Pro ($600/year). The technology is especially valuable for game development, e-commerce, AR/VR content, and architectural visualization.
Features
- ●Image to 3D: Generate 3D models from single images
- ●Text to 3D: Create 3D objects from text descriptions
- ●World Generation: Build complete 3D environments
- ●Texture & Materials: AI-inferred textures and PBR materials
- ●Multi-Format Export: Export to GLB, OBJ, and other 3D formats
Pricing
Tinkerer
$0
- ~10 free generations
- Core features
- Public designs
Starter
$50
/ year
- Monthly credits
- Private content
- Priority support
Maker
$200
/ year
- More credits
- All features
- Commercial use
Creative Pro
$600
/ year
- Maximum credits
- Priority processing
- API access
Pros & Cons
Pros
✓ Single-image 3D generation is fast and accessible
✓ Google DeepMind acquisition signals strong future
✓ Affordable yearly pricing for indie creators
✓ Supports multiple 3D export formats
Cons
✗ Pricing may change post-Google acquisition
✗ Quality varies on complex objects
✗ Limited community compared to established 3D tools
✗ Credit system can be confusing
Tags:3D generation, image to 3D, AI 3D, Google DeepMind, game assets