Nemotron 3.5 Super VL price cut on OpenRouter, now available on SageMaker JumpStart

NVIDIA Nemotron 3.5 Super VL now costs $0.15 per million input tokens and $0.47 per million output tokens on OpenRouter, down from $0.50 per million output tokens the day before. Input pricing is unchanged. The model has a 1 million token context window.

AWS Machine Learning announced NVIDIA Nemotron 3.5 Super VL is available in Amazon SageMaker JumpStart. The model accepts images, video, and text as input and generates text. It pairs a 120B-parameter hybrid Mixture-of-Experts language backbone with 12B active parameters and a C-RADIO vision encoder. It supports context lengths of up to 1 million tokens and includes a built-in Multi-Token Prediction head that supports speculative decoding. JumpStart offers two variants: NVFP4 and BF16 full-precision reference weights. NVFP4 checkpoint size is about 75 GB and BF16 checkpoint size is about 249 GB. NVFP4 runs with 4-bit weights and activations on Blackwell GPUs, on ml.g7.24xlarge and ml.g7.48xlarge instances. BF16 runs on ml.p5en.48xlarge instance. Deployment takes about 10 minutes on ml.g7.24xlarge.