Today we’re releasing OlmoEarth v1.1. It’s 3x cheaper to run than v1 while delivering the same state-of-the-art performance—and fully open. 🧵
Compute is the highest cost when running OlmoEarth at hundreds of thousands of square kilometers. Partners use v1 today for mangrove tracking, forest-loss classification, & country-scale crop-type mapping.
v1.1 makes that work more sustainable.
Where the savings come from: we feed the model about 3x fewer tokens per Sentinel-2 input.
Since compute scales quadratically with token count, even modest reductions compound into real efficiency gains.
Done naively, this hurts accuracy noticeably. Recovering it took changes to how we pretrain the model.
→ Full details in our tech report: https://t.co/wwCYuMGZ2I
token — The chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.
pretraining — The first, biggest phase of building a model: training it on enormous amounts of text so it learns language, facts, and reasoning in general.
Why it matters
OlmoEarth v1.1 runs about 3x cheaper at the same performance by feeding roughly 3x fewer tokenThe chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.Full definition → per Sentinel-2 input, with pretrainingThe first, biggest phase of building a model: training it on enormous amounts of text so it learns language, facts, and reasoning in general.Full definition → changes recovering the accuracy that naive token cuts lose. Holding data fixed from v1 isolates the effect.
Read the full thread on X ↗
