Olmo-3.1-32B-Instruct

A 32B instruction-tuned model from Ai2, built for reasoning, coding, and instruction-following, and released as a fully open model with public code, checkpoints, and training data.