Ai2 releases OlmoEarth v1.3 with precomputed global satellite embeddings

Ai2 released OlmoEarth v1.3, a satellite embedding model that reads a year of Sentinel-1 radar and Sentinel-2 and Landsat multispectral time series and produces a 128-dimensional vector per 10-meter pixel. Precomputed annual global embeddings are shared on Source Cooperative, along with the browser-based OlmoEarth Embeddings Explorer. The company reports the embeddings score highest on 20 of 27 evaluations against AlphaEarth Foundations and Tessera v2, and that v1.3 needed 53% of the compute of v1.2. Weights, pretraining data, and code are released openly.

Key takeaways

• OlmoEarth v1.3 produces 128-dimensional embeddings per 10m pixel, with the first 64 dimensions usable as a smaller embedding.
• Across 27 evaluations, the 128-dimensional embeddings score highest on 20 of 27 against AlphaEarth Foundations and Tessera v2.
• Training v1.3 used 53% of the computing time of v1.2 and 18% of v1, per the company's report.
• Precomputed global embeddings for 2025, 2024, and 2023 are available, with 2017 to 2022 covering select countries.

Why it matters

Precomputed global embeddings let users run lightweight classifiers locally without GPU access or storing raw imagery, lowering the barrier to land-cover, crop, and change-detection analysis. Openly released weights, data, and code allow independent reproduction.

Read with care

Benchmark results are the company's own evaluations, including a new benchmark it introduced (PLANTEUR); source claims not independently checked.