Prithvi WxC
NASA IMPACTPrithvi WxC is a NASA-IBM weather and climate foundation model trained on 160 MERRA-2 variables. Its 2.3-billion-parameter architecture combines masked reconstruction with forecasting, accepts flexible time intervals, and has separate base and rollout checkpoints plus downstream adaptations for regional forecasting and climate processes.
Openness
3 high confidence- license
- CDLA-Permissive-2.0(permissive non-OSI)
- weights
- open(CDLA-Permissive-2.0 ungated checkpoint)
- data
- documented-not-released(MERRA-2 variables and climatology identified
- code
- partial(model, validation, and downstream task code without full pretraining pipeline)
The downloadable checkpoint has a permissive data-oriented license and no use ceiling; the missing pretraining pipeline keeps it at the open-weights center.
- https://huggingface.co/ibm-nasa-geospatial/Prithvi-WxC-1.0-2300M/raw/main/README.md recorded 2026-10-21
Declares CDLA-Permissive-2.0, a 2.3B model trained on 160 MERRA-2 variables, and two pretrained checkpoint variants.
- https://raw.githubusercontent.com/NASA-IMPACT/Prithvi-WxC/main/README.md recorded 2026-10-21
Identifies the repository as model and downstream-task code, links task-specific fine-tuning repositories, and documents MERRA-2 climatology.
Adoption
1 high confidenceDownloads of the declared checkpoint on Hugging Face are the adoption signal here.
- https://huggingface.co/api/models/ibm-nasa-geospatial/Prithvi-WxC-1.0-2300M recorded 2026-09-10
downloads: 2138; gated: false; private: false.
Capability
3 high confidencePrithvi WxC documents two readable capabilities: task-specific downstream adaptation and a variable regional-to-global, flexible-lead-time forecasting scale. Producing a future timestamp from two input timestamps is a deterministic forecast rather than a generated output, and the 160 MERRA-2 variables it ingests count as one input system, not several.
- https://raw.githubusercontent.com/NASA-IMPACT/Prithvi-WxC/main/README.md recorded 2026-09-10
Documents downstream fine-tuning applications, masked reconstruction and forecasting, flexible lead times, and global and regional use.
Verified 2026-09-10