Open sourceNemotron 3.5 Super VL: NVIDIA releases an open multimodal model
NVIDIA has released Nemotron 3.5 Super VL, an open 120B-parameter model that reads images and PDFs, along with the smaller Bolt 74B-A9B for local use.
Source of truthblogs.nvidia.com/…/nemotron-3-5-super-vlOn October 15, 2026, NVIDIA released Nemotron 3.5 Super VL, with 120B parameters (12B active) and context of up to 1M tokens. It reads images, webpages and PDFs, including tables and charts. It ships with open weights, open datasets and training recipes, on Hugging Face, OpenRouter and as NVIDIA NIM. Alongside it came Bolt 74B-A9B, which uses 2-bit quantization to fit in 24GB of memory for local use; Bolt E4B is announced for October 23. The Hugging Face model cards list the OpenMDW License v1.1. NVIDIA says Super VL is up to 5x faster than other models in its class.
Other sources: huggingface.co/…/NVIDIA-Nemotron-3.5-Su…huggingface.co/…/NVIDIA-Nemotron-3.5-Bo…
Related stories
Granite Speech 5.0: IBM releases Apache 2.0 speech models
IBM released granite-speech-5.0-2b and revised two more speech models, all under the Apache 2.0 license. Greek is not listed for the 2B and 470M models.
Beam: Reflection AI’s open 501B model under Apache 2.0
Reflection AI has published the weights of Beam, a 501B-parameter model, on Hugging Face under Apache 2.0, which also allows commercial use.