ModelNemotron Bolt E4B
Nemotron 3.5 Bolt E4B
A lightweight open-source model released by NVIDIA with a memory footprint of approximately 4GB, aimed at local AI applications in edge scenarios such as phones, thin-and-light laptops and embedded devices.
Timeline (last 90 days)
Oct 23
NVIDIA fine-tuned Nemotron 3.5 Bolt E4B based on Gemma 4 E4B, surpassing the original Gemma 4 E4B on 16 of 21 NVIDIA tests.
Unverified50%
Oct 23
Nemotron 3.5 Bolt E4B scores 63.2 on Tau2-Bench, while the original Gemma 4 E4B scores 38.7.
Unverified50%
Oct 14
The same source claims that NVIDIA Nemotron 3.5 Bolt E4B will be released on October 23, with a memory footprint of about 4GB.
Expired50%