Understand this story
On November 10, 2026, AWS announced that NVIDIA Nemotron 3.5 Super VL is now available on Amazon SageMaker JumpStart, and users can deploy the model through SageMaker Studio or the Python SDK. According to information published by AWS Machine Learning, this vision-language model uses a 120B-parameter hybrid MoE language backbone, activating 12B parameters per token, paired with a C-RADIO vision encoder, supports a context length of up to 1 million tokens, and offers an NVFP4 quantized version and a BF16 full-precision version. On MMLongBench-Doc, the NVFP4 version scores the same as the BF16 version, both at 62.82, with gaps on the remaining multimodal benchmarks within about 1 point. There is currently no information on further developments.
AI-generated from public reports
Latest development · Report
NVIDIA Nemotron 3.5 Super VL Now Available on Amazon SageMaker JumpStartDevelopments
1 confirmed developments
NVIDIA Nemotron 3.5 Super VL is now available on SageMaker JumpStart
Source:AWS Machine Learning
Reports and discussion timeline
Browse reports and discussion by time. Their evidence levels remain distinct.
Nov 10
AWS Machine LearningFirst-party
NVIDIA Nemotron 3.5 Super VL Now Available on Amazon SageMaker JumpStart
NVIDIA Nemotron 3.5 Super VL is now available on Amazon SageMaker JumpStart and can be deployed via SageMaker Studio or the Python SDK. This vision-language model uses a 120B-parameter hybrid MoE language backbone with 12B parameters activated per token, paired with a C-RADIO vision encoder, supports a context length of up to 1 million tokens, and is offered in an NVFP4 quantized version and a BF16 full-precision version. NVFP4 matches BF16 on MMLongBench-Doc (62.82), with a gap of about 1 point or less on the other multimodal benchmarks.