NVIDIA

nvidia-ising-calibration-1.5-31b

Container

NVIDIA

nvidia-ising-calibration-1.5-31b

NVIDIA Ising Calibration 1.5 31B NIM provides BF16 and NVFP4 multimodal models for analyzing quantum-computing calibration experiment plots through an OpenAI-compatible API.

NVIDIA NIM

NVIDIA Ising Calibration 1.5 31B NIM Overview

Description

The NVIDIA Ising Calibration 1.5 31B NIM houses the NVIDIA-Ising-Calibration-1.5-31B-BF16 and NVIDIA-Ising-Calibration-1.5-31B-NVFP4 models, which are purpose-built dense multimodal vision-language models built on Gemma 4 31B, specialized for analyzing quantum computing calibration experiment plots. The models accept calibration experiment plot images and generate structured analytical outputs including technical descriptions, experimental conclusions, significance assessments, fit quality evaluations, parameter extractions, and experiment success classifications.

The container components are ready for commercial/non-commercial use.

Governing Download Terms

GOVERNING DOWNLOAD TERMS: Use of this container is governed by the NVIDIA Software License Agreement and Product-Specific Terms for NVIDIA AI Products. Use of this model is governed by the OpenMDW License Agreement, version 1.1. ADDITIONAL INFORMATION: Apache License, Version 2.0.

You are responsible for ensuring that your use of NVIDIA provided models complies with all applicable laws.

Deployment Geography: Global

Release Date:

Build.Nvidia.com [07/23/2026] via build.nvidia.com
NGC [07/23/2026] via NGC

Program Classes:

The NVIDIA Ising Calibration 1.5 31B NIM Container includes the following models:

Model Name	Use Case	How to Pull the Model
bf16 and NVFP4	Quantum calibration experiment image classification and analysis	Automatic

Deployment Details:

The NIM is deployed as a Docker container exposing an OpenAI-compatible HTTP API on port 8000 for chat completion and multimodal image analysis. Tool calling is not supported.

Our AI models are designed and/or optimized to run on NVIDIA GPU-accelerated systems. By leveraging NVIDIA's hardware (e.g. GPU cores) and software frameworks (e.g., CUDA libraries), the model achieves faster training and inference times compared to CPU-only solutions.

Reference(s):

Gemma 4 31B
QCalEval benchmark dataset for quantum calibration experiment evaluation
vLLM inference engine for high-throughput serving

Container Version(s):

Version	Precision	Description
nvidia/NVIDIA-Ising-Calibration-1.5-31B:1.5.0	BF16 and NVFP4	Production deployment with BF16 and NVFP4 model profiles

Software Stack:

Component	Version
NIM Version	1.5.0
Base Image	nvcr.io/nim/google/gemma-4-31b-it:1.7.1-variant
Backend	vLLM
CUDA	12.9.1
PyTorch	2.11.0+cu129

Model Specifications:

Property	Value
Architecture	Gemma 4 31B dense multimodal model
Total Parameters	~31B
Precision	BF16 (bfloat16) and NVFP4
Model Maximum Context Length	262,144 tokens
NIM Default Context Length	128,000 tokens
Tensor Parallelism	BF16: 1 or 2; NVFP4: 1
Pipeline Parallelism	1

Hardware Support Matrix

Hardware systems that have undergone QA testing:

GPU	GPU Memory (GB)	Precision	# of GPUs	Disk Space (GB)
L40S	48	BF16 and NVFP4	2	72
H200 SXM	141	BF16 and NVFP4	1 or 2	72
GH200	96/144	BF16 and NVFP4	1 or 2	72
B200	192	BF16 and NVFP4	1 or 2	72
GB200	192	BF16 and NVFP4	1 or 2	72
GB300	288	BF16 and NVFP4	1	72
DGX Spark	128	BF16 and NVFP4	1	72

Requires NVIDIA Ada Lovelace, Hopper, or Blackwell architecture.

Security Common Vulnerabilities and Exposures (CVEs)

Please review the Security Scanning tab on NGC to view the latest security scan results. For certain open-source vulnerabilities listed in the scan results, NVIDIA provides a response in the form of a Vulnerability Exploitability eXchange (VEX) document. The VEX information can be reviewed and downloaded from the Security Scanning tab.

Ethical Considerations:

NVIDIA believes Trustworthy AI is a shared responsibility and we have established policies and practices to enable development for a wide array of AI applications. Developers should work with their internal developer team to ensure these software components meet requirements for the relevant industry and use case and address unforeseen product misuse.

Please report quality, risk, security vulnerabilities or NVIDIA AI Concerns here.

Get Help

Getting started with the NIM

Deploying and integrating the NIM is straightforward thanks to our industry standard APIs. Visit the NIM Container page for release documentation, deployment guides and more.