Skip to content

OLMo

An open language model with a full training framework

Visit official site

What OLMo is

The framework is described as an open language model and training framework released by AI2, giving researchers the full set of training data, code, model parameters and evaluation code. The material says this openness lowers the barrier to language model research.

What you can do with it

  • Train and experiment with large language models
  • Read the training data, code and evaluation code
  • Use the documentation for open research work

Application scenarios

  • Train a language model at 1 billion parameters
  • Carry out language understanding research on its models
  • Adjust some hyperparameters from the training method

Pros & cons

✓ What we like

  • Full information from data to model is accessible
  • The stack supports open research and collaboration

! What to watch out for

  • The release is aimed at research and experimentation
  • Model work depends on the provided framework

FAQ

Who released it?

The material states it was released by the AI2 research institute.

What resources are included?

Training data, code, model parameters and evaluation code are listed.

Last reviewed: 2026-11-15

More LLM API platform tools

View all →

How we review