{"slug": "ai2-releases-satellite-embeddings-that-let-mapping-teams-skip-repeat-model-runs", "title": "Ai2 releases satellite embeddings that let mapping teams skip repeat model runs", "summary": "Ai2 released OlmoEarth v1.3 on October 22nd, pairing a satellite model with precomputed annual 10-meter embeddings so mapping teams can build new maps without rerunning the model over imagery for each task. Coverage is global for 2024 and 2025, near-global for 2023, and limited to selected countries for 2017-2022, and Ai2 says its embeddings led two named competitors on most of its evaluations. The Seattle nonprofit, founded in 2014 by late Microsoft co-founder Paul Allen, makes the features publicly accessible, though uneven year-by-year coverage and artifact license restrictions shape where teams can use them.", "body_md": "# Ai2 releases satellite embeddings that let mapping teams skip repeat model runs\n\n**OlmoEarth v1.3 supplies precomputed 10-meter features and a browser-based mapping tool. Coverage is global for 2024 and 2025, near-global for 2023, and limited to selected countries for 2017-2022; Ai2 says its embeddings led two named competitors on most of its evaluations.**\n\n        By [RuntimeWire Staff](https://runtimewire.com/author/runtimewire-staff)\n        · Published \n\nPrimary source: [Hugging Face Newsroom](https://huggingface.co/blog/allenai/olmoearth-v1-3)\n\n## Why it matters\n\nPrecomputed features let mapping teams test many downstream questions without rerunning a large satellite model for each one. OlmoEarth makes that data publicly accessible, but the uneven year-by-year coverage and artifact license restrictions shape where and how teams can use it.\n\n[Ai2](https://allenai.org/?ref=runtimewire), the Seattle nonprofit founded in 2014 by late Microsoft co-founder Paul Allen, released [OlmoEarth v1.3](https://huggingface.co/blog/allenai/olmoearth-v1-3?ref=runtimewire) on October 22nd. The release pairs a satellite model with precomputed annual embeddings: reusable features intended to let researchers make new maps without running the model over satellite imagery again for each task.…\n\nWhen you purchase through links in our articles, [we may earn a small commission](https://runtimewire.com/affiliate-monetization-standards). This doesn’t affect our editorial independence.", "url": "https://wpnews.pro/news/ai2-releases-satellite-embeddings-that-let-mapping-teams-skip-repeat-model-runs", "canonical_source": "https://runtimewire.com/article/ai2-olmoearth-v1-3-satellite-embeddings", "published_at": "2026-10-23 14:16:00+00:00", "updated_at": "2026-10-23 17:40:45.911023+00:00", "lang": "en", "topics": ["machine-learning", "computer-vision", "ai-research"], "entities": ["Ai2", "OlmoEarth v1.3", "Paul Allen", "Hugging Face", "Microsoft"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/ai2-releases-satellite-embeddings-that-let-mapping-teams-skip-repeat-model-runs", "markdown": "https://wpnews.pro/news/ai2-releases-satellite-embeddings-that-let-mapping-teams-skip-repeat-model-runs.md", "text": "https://wpnews.pro/news/ai2-releases-satellite-embeddings-that-let-mapping-teams-skip-repeat-model-runs.txt", "jsonld": "https://wpnews.pro/news/ai2-releases-satellite-embeddings-that-let-mapping-teams-skip-repeat-model-runs.jsonld"}}