aws AWS What's New ·

New GLM-5.2 FP8, Nemotron-Nano-12B-v2, and GLM-OCR models on SageMaker JumpStart

aiawsgaengineeraws-sagemaker
feature announcement

Amazon SageMaker JumpStart now offers Z.ai's GLM-5.2 FP8, NVIDIA's Nemotron-Nano-12B-v2, and Z.ai's GLM-OCR models. These additions expand the portfolio of foundation models, providing specialized capabilities for long-horizon agentic engineering, efficient hybrid reasoning, and advanced document understanding. AWS customers can leverage these models to deploy high-performance, scalable AI solutions for tasks like full-cycle software development, enterprise applications, and large-scale document processing. Deployment is streamlined via the SageMaker console or SDK, enabling rapid integration of advanced AI into existing workflows.

  • GLM-5.2 FP8 for Long-Horizon Agentic Engineering
  • NVIDIA Nemotron-Nano-12B-v2 for Efficient Hybrid Reasoning
  • GLM-OCR for Advanced Document Understanding
Features (3)
  • GLM-5.2 FP8 for Long-Horizon Agentic Engineering

    Optimized for long-horizon tasks and agentic engineering workflows, this model offers a truly usable 1M-token context window. It enables reliable execution of long-running tasks and consistent adherence to engineering standards for full development workflows from requirements to deployment.

  • NVIDIA Nemotron-Nano-12B-v2 for Efficient Hybrid Reasoning

    This model excels in unified reasoning and non-reasoning tasks with high inference throughput, using a hybrid Mamba-2 and Transformer architecture with a 128K context length. Its compact 12B parameter design achieves comparable or better accuracy than leading open models while delivering up to 6x higher inference throughput.

  • GLM-OCR for Advanced Document Understanding

    A 0.9B-parameter multimodal model, GLM-OCR provides accurate, fast, and comprehensive understanding for complex real-world documents including scanned PDFs, handwritten notes, and academic papers. It reconstructs structure, tables, and formulas into clean Markdown, JSON, or LaTeX, with latency low enough for real-time services and edge devices.

Read the original announcement →

https://aws.amazon.com/about-aws/whats-new/2026/01/glm-5.2-fp8-nemotron-nano-12b-v2-glm-ocr-on-sagemaker-jumpstart/

Related releases