How an Agent Built a 3D Paris Gallery by Chaining Two Hugging Face Spaces

A recent blog post on Hugging Face demonstrates how an AI agent can chain two existing Hugging Face Spaces to produce a 3D virtual gallery of Paris landmarks. One Space generates 3D models from text prompts, while the other assembles them into an interactive gallery. The agent orchestrates the pipeline by passing outputs from the first Space as inputs to the second, showing a practical pattern for composing modular AI services. This approach, according to the Hugging Face Blog, lowers the barrier for creating complex multimodal experiences without writing extensive code. For developers building AI workflows, the example highlights the value of reusable components and the potential of agent-driven orchestration. Rather than training new models, one can leverage existing ones connected via lightweight agents. The post also underscores the growing ecosystem of hosted AI spaces that can be combined on the fly.