Nemotron 3.5 Lightning vs MiMo-V2.6-Pro

Pyyan / Compare / Nemotron 3.5 Lightning vs MiMo-V2.6-Pro

2 of 5

Open-Weight Models · verified 29 Sept 2026

Nemotron 3.5 LightningNVIDIAcurrent

XMiMo-V2.6-ProXiaomicurrent

Add another3 slots left

Specification	Nemotron 3.5 Lightning	MiMo-V2.6-Pro
Summary	Hybrid Mamba and transformer, 3B active, and the training data ships with it.	A trillion parameters under the MIT licence.
Context	128K	1M
Input ($/Mtok)	Weights only	Weights only
Output ($/Mtok)	Weights only	Weights only
Licence	OpenMDW-1.1	MIT
Max output	32K	Not published
Parameters	30B total, 3B active	1.02T total, 42B active
Released	11 Aug 2026	21 Sep 2026
Model ID	nvidia/nemotron-3.5-lightning	XiaomiMiMo/MiMo-V2.6-Pro
Category	Open-Weight Models	Open-Weight Models
Official	—	Xiaomi ↗

Highlighted rows are where these differ.

30B mixture of experts with roughly 3B active per token, hybrid Mamba and transformer
Weights, training data and recipes all released, under NVIDIA's OpenMDW-1.1 licence
Scores 24 on the Artificial Analysis Intelligence Index, up from 15 for Nemotron 3 Nano

Best for long-running agents.

1.02T total parameters, 42B active per token, a 4.1% activation ratio
Text, image, video and audio in, with a 1M token context
Shipped with 7,000+ reinforcement learning environments and the framework that trained it

Best for multimodal work on your own cluster.