The OlmoEarth Platform: Geospatial inference at planetary scale
The Allen Institute for AI (AI2) introduced the OlmoEarth Platform, a new infrastructure for geospatial inference at planetary scale. The platform is built to handle massive geospatial datasets, enabling AI models to perform tasks such as land cover classification, change detection, and object detection across the entire Earth. While specific model names or versions were not detailed in the announcement, the platform is designed to support a range of geospatial AI workloads. The OlmoEarth Platform is hosted on Hugging Face, making it accessible to the broader AI and geospatial communities. This release represents a significant step in scaling geospatial AI, potentially reducing the barriers for researchers and developers to conduct planetary-scale analysis. The platform's infrastructure is optimized for efficiency, allowing users to run inference on large-scale geospatial data without requiring extensive computational resources. The announcement emphasizes the platform's role in enabling new applications in climate science, agriculture, and disaster response, though specific use cases or performance benchmarks were not provided.
Developers gain access to a scalable platform for planetary-scale geospatial AI inference.