Ai2 introduces Molmo2-ER as the MolmoAct2 reasoning backbone
The May5 Ai2 post identifies Molmo2-ER, a separately specialized vision-language backbone for spatial questions, pointing and image/video reasoning that feeds MolmoAct2’s action model. It reports a63.8/100 mean across13 embodied reasoning benchmarks. This is an introduction date, not a verified first-public-weight day.
Previously reported
Molmo2 supplies the general vision-language starting point before embodied specialization. This describes a research baseline, not an earlier robot controller.
This announcement
The May5 Ai2 post identifies Molmo2-ER, a separately specialized vision-language backbone for spatial questions, pointing and image/video reasoning that feeds MolmoAct2’s action model. It reports a63.8/100 mean across13 embodied reasoning benchmarks. This is an introduction date, not a verified first-public-weight day.
Ordinary historical component introduction: the same May5 MolmoAct2 launch already has a featured event. Preserve this distinct model role without a duplicate featured launch headline. Admission94 does not determine importance; no separate first code/weight/API opening is dated here.