aws AWS What's New ·

SageMaker JumpStart adds NVIDIA Qwen3.6-35B-A3B-NVFP4 and Alibaba Wan2.1-T2V-1.3B models

aiawsgaengineeraws-sagemaker
feature

Amazon SageMaker JumpStart now provides access to NVIDIA's Qwen3.6-35B-A3B-NVFP4 and Alibaba's Wan2.1-T2V-1.3B-Diffusers models. These additions expand the foundation model portfolio, enabling customers to deploy specialized AI solutions on AWS infrastructure. Qwen3.6 offers agentic coding and long-context reasoning with a reduced memory footprint through NVFP4 quantization. Wan2.1 provides efficient text-to-video generation on consumer-grade hardware, making it highly accessible. Customers can deploy these models via the SageMaker console or Python SDK.

  • NVIDIA Qwen3.6-35B-A3B-NVFP4 for Agentic Coding
  • Alibaba Wan2.1-T2V-1.3B-Diffusers for Text-to-Video Generation
Features (2)
  • NVIDIA Qwen3.6-35B-A3B-NVFP4 for Agentic Coding

    This NVIDIA-quantized variant of Alibaba's Qwen3.6-35B-A3B is optimized for agentic coding, multimodal reasoning, and long-context understanding, supporting up to ~1M tokens. The 35B-parameter Mixture-of-Experts model, quantized to NVFP4 using NVIDIA's ModelOpt, supports multi-step agent pipelines with a significantly reduced memory footprint.

  • Alibaba Wan2.1-T2V-1.3B-Diffusers for Text-to-Video Generation

    Built on the diffusion transformer paradigm with a novel Video Variational Autoencoder, this 1.3B-parameter model excels at generating high-quality, physics-consistent video clips from text prompts. It can produce a 5-second 480p video on an RTX 4090 in approximately 4 minutes, requiring only 8.19 GB of VRAM.

Read the original announcement →

https://aws.amazon.com/about-aws/whats-new/2026/01/qwen3.6-35b-a3b-nvfp4-wan2.1-t2v-1.3B-diffusers-jumpstart/

Related releases