country_code
Skip to main content
Ctrl+K
NVIDIA TensorRT for RTX - Home NVIDIA TensorRT for RTX - Home

NVIDIA TensorRT for RTX

  • Documentation Home
NVIDIA TensorRT for RTX - Home NVIDIA TensorRT for RTX - Home

NVIDIA TensorRT for RTX

  • Documentation Home

Table of Contents

Getting Started

  • Release Notes
    • 1.5 (Latest)
    • 1.4
    • 1.3
    • 1.2
    • 1.1
    • 1.0
  • Support Matrix
    • 1.5 (Latest)
    • 1.4
    • 1.3
    • 1.2
    • 1.1
    • 1.0

Installing TensorRT-RTX

  • Installation Guide Overview
  • Prerequisites
  • Installing TensorRT-RTX
  • Deploy Your First Model
  • ONNX Conversion Guide
  • Using TensorRT-RTX via PyTorch

Architecture

  • Architecture Overview

Inference Library

  • Using the Native Runtime API
  • C++ API Documentation
  • Python API Documentation
  • Advanced Topics
  • Working with Quantized Types
  • Working with Dynamic Shapes
  • Working with Runtime Cache
  • Working with RTX CUDA Graphs
  • Simultaneous Compute and Graphics
  • CPU-Only AOT and TensorRT-RTX Engines
  • Porting Guide for TensorRT Applications
  • TensorRT-RTX API Capture and Replay

Performance

  • Best Practices

API

  • C++ API
  • Python API
  • TensorRT-RTX Operators
  • ONNX GraphSurgeon API
  • Polygraphy API

Reference

  • Deprecation Policy
  • Cybersecurity Disclosures
  • NVIDIA SOFTWARE LICENSE AGREEMENT
  • ONNX GraphSurgeon API
Is this page helpful?

ONNX GraphSurgeon API#

TensorRT-RTX supports loading models from ONNX using the ONNX Parser library. ONNX GraphSurgeon provides a convenient way to create and modify ONNX models.

Interface - ONNX GraphSurgeon API

GitHub - TensorRT > tools > onnx-graphsurgeon

See also

Polygraphy API

Companion toolkit for running and debugging deep learning models.

Architecture Overview

ONNX import paths and model preparation workflows.

previous

TensorRT-RTX Operators

next

Polygraphy API

NVIDIA NVIDIA
Privacy Policy | Your Privacy Choices | Terms of Service | Accessibility | Corporate Policies | Product Security | Contact

Copyright © 2025-2026, NVIDIA Corporation.

Last updated on May 28, 2026.