AlphaEarth Foundations
GoogleAlphaEarth Foundations is Google DeepMind's embedding field model for Earth observation. It assimilates spatial, temporal, and measurement context from Landsat, Sentinel-1, Sentinel-2, and other sources into a 64-dimensional embedding for every 10-meter pixel of terrestrial land and shallow coastal water, published as annual Satellite Embedding layers in Google Earth Engine covering 2017 onward.
What Google DeepMind released is the embedding dataset, not the model. There is no checkpoint, no implementation, and no license over weights, so the product is scored closed even though its outputs are freely usable.
Openness
1 high confidence- license
- proprietary(no license granted over the model, the embedding layers are licensed separately under CC-BY 4.0)
- weights
- closed(no checkpoint distributed, only precomputed annual embedding layers)
- data
- closed(training corpus not released)
- code
- closed(no public implementation, training, or inference code)
Both primary sources describe releasing a dataset of embedding layers, not the model itself. Because no weights are downloadable, the license question does not even arise, and the CC-BY-4.0 grant on the Earth Engine collection covers the output layers, not the model that produced them.
- https://arxiv.org/abs/2507.22291 recorded 2026-10-21
Abstract states that the authors have released a dataset of global, annual, analysis-ready embedding field layers from 2017 through 2024, and offers no weights, corpus, or code.
- https://deepmind.google/discover/blog/alphaearth-foundations-helps-map-our-planet-in-unprecedented-detail/ recorded 2026-10-21
Announcement says the team is releasing a collection of AlphaEarth Foundations' annual embeddings as the Satellite Embedding dataset in Google Earth Engine, with no mention of publishing the model.
- https://storage.googleapis.com/earthengine-stac/catalog/GOOGLE/GOOGLE_SATELLITE_EMBEDDING_V1_ANNUAL.json recorded 2026-10-21
The collection's STAC record distributes an Earth Engine ImageCollection of 64-band annual embedding images under CC-BY-4.0 with a required attribution string. It describes a dataset; nothing in it distributes the model. Cited in place of the rendered catalog page, whose body is nondeterministic.
Adoption
1 medium confidenceNo model artifact is distributed, so there is no download or star count to read and no reach word is recorded. DeepMind reports working with more than 50 organizations on the released embedding dataset, which is directional only.
- https://deepmind.google/discover/blog/alphaearth-foundations-helps-map-our-planet-in-unprecedented-detail/ recorded 2026-09-10
States that over the past year the team worked with more than 50 organizations to test the dataset on real-world applications.
Capability
2 high confidenceAlphaEarth Foundations combines Landsat, Sentinel-1 and Sentinel-2 measurements into one embedding, which is its only counted capability. The embeddings are applied without retraining rather than adapted, an embedding is not a generated output, and the release fixes a single 10-meter annual grid rather than multiple scales.
- https://arxiv.org/abs/2507.22291 recorded 2026-09-10
Describes an embedding field model that assimilates spatial, temporal, and measurement contexts across multiple sources, and reports that its embeddings outperform other featurization approaches without re-training.
- https://storage.googleapis.com/earthengine-stac/catalog/GOOGLE/GOOGLE_SATELLITE_EMBEDDING_V1_ANNUAL.json recorded 2026-09-10
The collection's STAC record — a single 10-metre pixel size, one 64-dimensional embedding per pixel per calendar year, and the landsat-derived, sentinel1-derived and sentinel2-derived keywords. Cited in place of the rendered catalog page, whose body is nondeterministic and returned three different digests in three consecutive fetches, so it could never be re-confirmed.
Verified 2026-09-10