Model reference · open weights

maira-2

LLMs microsoft Text gen 1 build Its own licence terms 3k dl/mo

maira-2 is an open-weight language model from microsoft. maira-2 (FP32) weighs 13.8 GB; the smallest configuration that runs it is RTX 4060 Ti 16 GB.

MAIRA-2 is a multimodal transformer developed by Microsoft Research Health Futures for generating grounded or non-grounded radiology reports from chest X-rays.
The model contains 6.9B parameters and operates in English, utilizing a frozen RAD-DINO-MAIRA-2 image encoder and a fine-tuned vicuna-7b-v1.5 language model.
It is intended for research purposes only and is not meant for clinical practice, with a licence identified as MSRLA.

Summary of the microsoft/maira-2 model card, 2026-10-01

What it is

What it runs on

Memory and cards for maira-2 (FP32)

13.8 GBweights, file size

762 MBruntime overhead, at least

How much memory each request adds isn't estimated yet for this architecture. The weights need at least the cards below, plus room for the context.

How it works

Details