/ Models Index / OLMo 2

OLMo 2

Open-Weights · AI2 · Allen Institute for AI · Released Nov 2024
B-

Good (B-) — Capability Grade

OLMo 2 is the Allen Institute for AI's fully-open model — Apache 2.0 weights, open data, open training code. Strong for research reproducibility. Mid-tier capability.

74
Composite / 100
/ Subscore Breakdown · 6 Capability Dimensions

Where this grade comes from.

General Reasoning
B-
Code Generation
C+
Math & STEM
C+
Tool Use & Agency
C
Multimodal
D
Safety & Alignment
B
/ Key Events & Disclosures

Release timeline & positioning.

  • Released Nov 2024
  • Fully-open (weights + data + training code)
  • AI2 research lab
  • Strong reproducibility positioning

/ Best for

Research applications requiring full training-pipeline transparency.

/ Watch out for

Capability lags frontier open-weights peers. Reserve for research-reproducibility-required applications.