---
title: "Introducing OlmoEarth embeddings: Custom embedding exports from OlmoEarth Studio for downstream analysis"
description: "OlmoEarth Studio now supports custom computation and export of embedding vectors, compact representations of Earth-observation data generated by open-source OlmoEarth foundation models. Users can select an area, a one-to-twelve-month time span, an encoder variant, spatial resolution, and Sentinel-2 L2A or Sentinel-1 RTC imagery, then receive a Cloud-Optimized GeoTIFF with one int8 band per embedding dimension. The post demonstrates similarity search, few-shot segmentation, change detection, and PCA-based exploration; a logistic regression trained on 60 labeled pixels produced a weighted F1 of 0.84 for three land-cover classes in Ca Mau, Vietnam. Embeddings provide a fast, shareable starting point, while supervised fine-tuning can improve performance for applications needing task-specific results; cloud cover, atmospheric artifacts, and missing observations can affect quality."
---

# Introducing OlmoEarth embeddings: Custom embedding exports from OlmoEarth Studio for downstream analysis

[Hugging Face](https://yomu.fyi/company/hugging-face) · Kyle Wiggers · Aug 12, 2026

**Type:** Announcement

## Summary

OlmoEarth Studio now supports custom computation and export of embedding vectors, compact representations of Earth-observation data generated by open-source OlmoEarth foundation models. Users can select an area, a one-to-twelve-month time span, an encoder variant, spatial resolution, and Sentinel-2 L2A or Sentinel-1 RTC imagery, then receive a Cloud-Optimized GeoTIFF with one int8 band per embedding dimension. The post demonstrates similarity search, few-shot segmentation, change detection, and PCA-based exploration; a logistic regression trained on 60 labeled pixels produced a weighted F1 of 0.84 for three land-cover classes in Ca Mau, Vietnam. Embeddings provide a fast, shareable starting point, while supervised fine-tuning can improve performance for applications needing task-specific results; cloud cover, atmospheric artifacts, and missing observations can affect quality.

## Context

OlmoEarth embeddings are presented as a fast, cost-effective entry point for using Earth-observation foundation models in downstream analysis. The post motivates custom, on-demand generation so embeddings reflect selected locations, time periods, sensors, and seasonal conditions rather than relying only on a precomputed global archive.

## Approach / What changed

OlmoEarth Studio computes embeddings on demand after users configure an area of interest, time span, encoder variant, spatial resolution, and imagery sources through the UI or API. Results are exported as Cloud-Optimized GeoTIFFs with one signed 8-bit band per embedding dimension; Studio also supports supervised fine-tuning for task-specific applications.

## Takeaways

- Studio offers Nano (128-dimensional), Tiny (192-dimensional), and Base (768-dimensional) encoder variants, with 10 m, 20 m, 40 m, or 80 m spatial resolutions and Sentinel-2 L2A, Sentinel-1 RTC, or combined imagery.
- A logistic regression using standardized frozen embeddings and only 60 labeled pixels produced a coherent three-class land-cover map for Ca Mau, Vietnam, with weighted F1 of 0.84.
- Embedding quality depends partly on input imagery; persistent cloud cover, atmospheric artifacts, and missing observations during the composite period can affect resulting vectors.

**Tags:** [Machine Learning](https://yomu.fyi/topic/machine-learning), [Open Source](https://yomu.fyi/topic/open-source), [Performance](https://yomu.fyi/topic/performance), [Search](https://yomu.fyi/topic/search)

- Source: [Hugging Face](https://huggingface.co/blog/allenai/olmoearth-embeddings)
- Source URL: https://huggingface.co/blog/allenai/olmoearth-embeddings
- Ingested by Yomu: 2026-08-29T05:41:10.165Z

[Read original post](https://huggingface.co/blog/allenai/olmoearth-embeddings)