CSM

• Image-to-3D model generation
• Text-to-3D world creation
• Sketch / hand-drawn input to 3D conversion
• AI retopology and mesh optimization
• PBR texture generation and retexturing
• Multiview / multi-angle image upload support
• Integration with Blender via MCP (Model Context Protocol)
• Chat-to-3D (conversational prompt → 3D)
• API access and scalable use
• Priority support, private content, and mentorship in higher tiers

CSM (Common Sense Machines) is a generative AI platform designed to reshape how creators and developers build 3D content. The core promise of CSM is to take everyday inputs — images, sketches, or textual descriptions — and convert them into rich, interactive, production-ready 3D models and worlds. Whether you are building for games, VR/AR, animation, simulation, or architectural prototyping, CSM’s tools streamline workflows and reduce the barrier to entry for 3D design.

At its heart lies the ability to generate 3D assets from flat sources. Upload a photograph, a sketch, or write a high level prompt and CSM will infer depth, geometry, textures, and structure to produce a usable mesh. The system also supports advanced enhancements such as retopology (reworking mesh for optimal structure), PBR texture generation, and multi-view inputs to refine realism. Another standout is the integration with Blender via its Model Context Protocol (MCP) — enabling seamless transfer and use of CSM assets inside a professional 3D pipeline. Through this, creators can not only import models but also animate or interact with them via text or prompt-based commands.

CSM’s pricing uses a credit-based subscription model. A free “Tinkerer” tier gives users limited credits (roughly 10 generations) and access to core features on public servers. Paid tiers (“Maker”, “Creative Pro”, “Prime”, or Enterprise) expand credits, unlock private ownership of content, priority support, mentorship, and additional tools. This model allows hobbyists and professionals both to adopt CSM according to scale and budget.

Beyond content generation, CSM is building towards generative agents — AI systems that can learn workflows, replicate user actions, and handle complex editing tasks autonomously. Their blog discusses evolving agentic control over workflows like retopology, UV mapping, and PBR refinement. As this evolves, CSM aims to become not just a content tool but a creative copilot for 3D workflows.

In sum, CSM offers an ambitious and modern platform for turning simple inputs into immersive 3D outputs, bridging imagination and interactive worlds for creators of all levels.