NVIDIA’s Nemotron 3.5 Super VL, a 120-billion-parameter multimodal vision-language model, is now available through Amazon SageMaker JumpStart. According to the announcement, users can deploy the model in a few steps and choose between NVFP4 or BF16 variants. Once deployed, the model supports inference across text, image, and video inputs.
The stated use cases include document intelligence, computer-use agents, and broader multimodal agentic workloads.
Why it matters
Making a large multimodal model available through a managed deployment path lowers the setup effort for teams that want to run text, image, and video inference within an existing cloud environment.
Who should care
Developers and enterprises building document processing, computer-use agents, or multimodal agentic applications on AWS infrastructure.