Olmo 3

2025/12/15 by Team OLMo, Team Olmo, Olmo, Team +141 · 1 voice
Computer Science · #Multimodal Machine Learning Applications #Natural Language Processing Techniques #Topic Modeling #cs.CL #cs.LG

paper · pdf · doi:10.48550/arxiv.2512.13961

arxiv published 2025/12/15 · arxiv updated 2026/04/14

Abstract

We introduce Olmo 3, a family of state-of-the-art, fully-open language models at the 7B and 32B parameter scales. Olmo 3 model construction targets long-context reasoning, function calling, coding, instruction following, general chat, and knowledge recall. This release includes the entire model flow, i.e., the full lifecycle of the family of models, including every stage, checkpoint, data point, and dependency used to build it. Our flagship model, Olmo 3 Think 32B, is the strongest fully-open thinking model released to-date.

Discussions

• the data provenance and mix depends highly on the training stage / goals you have for your model. the OLMO 3 technical report is a great resource that goes into lots of detail: arxiv.org/pdf/2512.1396 [bsky, 3 points, 0 comments]

Related

• OLMo: Accelerating the Science of Language Models
• 2 OLMo 2 Furious
• OLMoE: Open Mixture-of-Experts Language Models
• OpenELM: An Efficient Language Model Family with Open Training and Inference Framework
• The Llama 3 Herd of Models