Skip to main content ->
Ai2

Latest research

November 20, 2025

Olmo 3: Charting a path through the model flow to lead open-source AI

Our new flagship Olmo 3 model family empowers the open source community with not only state-of-the-art open models, but the entire model flow and full traceability back to training data.
Read post
November 18, 2025

DR Tulu: An open, end-to-end training recipe for long-form deep research

We introduce Deep Research Tulu (DR Tulu), an open post-training recipe and framework for long-form deep research agents.
Read post
November 4, 2025

Introducing OlmoEarth Platform: Powerful open infrastructure for planetary insights

Introducing OlmoEarth Platform, an open, scalable, end-to-end system that turns multi-sensor Earth data into constantly-updating, decision-ready insights.
Read post
November 4, 2025

OlmoEarth: A new state-of-the-art Earth observation foundation model family

OlmoEarth is a new family of industry-leading open foundation models built to make Earth AI practical, scalable, and performant for real-world applications.
Read post
October 22, 2025

olmOCR 2: Unit test rewards for document OCR

olmOCR 2, our latest document OCR model, achieves state-of-the-art performance for English-language digitized print documents.
Read post
October 16, 2025

SamudrACE: Highly efficient coupled global climate modeling with the Ai2 climate emulator

SamudrACE couples 3D models of both the ocean and the atmosphere, giving it a deep understanding of global Earth system interactions.
Read post
October 8, 2025

Making AI citations count with Asta

We're releasing data that shows which scientific papers our agentic platform for research and discovery, Asta, relies on most when answering questions.
Read post
October 1, 2025

Asta DataVoyager: Data-driven discovery and analysis

DataVoyager is our new feature in Asta built to address the challenges scientists face in drilling down into structured datasets.
Read post
September 16, 2025

Fluid language model benchmarking

We explore how Fluid Benchmarking can adapt evaluation items to a language model’s capability level.
Read post