Organisation
Allen Institute for AI
A non-profit research institute known for fully open language models (OLMo) that publish training data, code and checkpoints, not just final weights.
The Allen Institute for AI, known as Ai2, is a Seattle non-profit research institute founded by the late Microsoft co-founder Paul Allen. In the language-model era it has pushed a stricter definition of "open" than most: with its OLMo models it publishes not just final weights but the full training data, code and intermediate checkpoints needed to reproduce and audit them — a deliberate contrast with weights-only releases such as Meta's Llama. Its releases are typically frontier-adjacent rather than frontier-leading, and it does not claim state-of-the-art performance; the aim is reproducibility and transparency for researchers. Through 2025 it continued that approach with Olmo 3 and tools such as OlmoTrace, which traces a model's outputs back to specific training documents.
- Category
- Open source & ecosystem
- Founded
- 2014
- HQ
- Seattle, US
- Key people
- Peter Clark
Appears alongside
Tracks
- Open weights & ecosystem 9
- Models & capabilities 6
- Benchmarks & progress 1
- Safety & alignment 1
- Ideas & essays 1
AI2 releases Olmo 3 open frontier model family
AI2 released Olmo 3 (7B, 32B), including a fully open 32B reasoning model, releasing every stage of the model flow from data to deployment.
Open weights & ecosystem · Models & capabilities
'The Leaderboard Illusion' paper critiques Chatbot Arena methodology
Researchers found Meta tested roughly 27 private Llama variants before its public release and that OpenAI and Google alone received about 40% of all Arena battle data.
Benchmarks & progress
AI2 releases OLMoTrace for tracing model outputs to training data
The tool highlights spans of a model's output that match its training data verbatim and links to the source documents, using an indexed search algorithm the researchers said scaled to trillions of tokens.
Open weights & ecosystem · Safety & alignment
AI2 releases OLMo 2
Trained on up to 5 trillion tokens, AI2 said the 7B and 13B models beat Llama 3.1 8B and Qwen 2.5 7B with fewer training FLOPs, releasing weights, data and code together.
Open weights & ecosystem · Models & capabilities
AI2 releases Tulu 3 post-training recipe
AI2 released the full data, code and recipe behind Tülu 3, noting that none of the top 50 models on the Chatbot Arena leaderboard had published their own post-training data.
Open weights & ecosystem · Models & capabilities
AI2 releases Molmo multimodal models
AI2 said its largest Molmo model trained on under a million image-text pairs, roughly three orders of magnitude less data than comparable systems, and released weights and data openly.
Open weights & ecosystem · Models & capabilities
AI2 releases OLMoE mixture-of-experts model
1 billion active of 7 billion total parameters, trained on 5 trillion tokens, released with 244 intermediate checkpoints and full training data and logs.
Open weights & ecosystem · Models & capabilities
AI2 releases OLMo, a fully open language model
Unlike other 'open' releases, AI2 published the full Dolma training corpus, training code and hundreds of intermediate checkpoints, not just final weights.
Open weights & ecosystem · Models & capabilities
AI2 releases Dolma open training corpus
The 3-trillion-token corpus was released under AI2's ImpACT licence, which requires users to register their intended use and disclose derivative works.
Open weights & ecosystem
The CORD-19 research dataset is released
Assembled with the White House OSTP, National Library of Medicine and Chan Zuckerberg Initiative, it opened with over 29,000 papers, most with full text.
Ideas & essays · Open weights & ecosystem
Also mentioned in 3 entries
Referenced in passing — Allen Institute for AI isn't the main subject of these.