MiMo-VL-7B
7B vision-language model with a native-resolution ViT, trained with Mixed On-policy RL across perception, grounding and reasoning rewards.
Four-stage pretraining then MORL; open SFT and RL checkpoints under MIT. A 2508 update followed.
- Date
- Friday, 30 May 2025
- Lab
- Xiaomi (MiMo)
- Kind
- open-weights
- Access
- open weights
Date is the HF weights upload (2025-05-30 UTC); Wikipedia lists 2025-06-04. No benchmark numbers read.
Sources
This record was checked against its sources on 6 October 2026. How we check