Built something? We create video reels & spotlights for GitHub projects.Promote your project →
pytorch
Home / Python / TensorRT

pytorch/TensorRT

PyTorch/TorchScript/FX compiler for NVIDIA GPUs using TensorRT

Python ◇ cuda BSD-3-Clause
★3KSTARS
⑂411FORKS
!342ISSUES
🏆#1,639GLOBAL RANK
🔥2DAYS TRENDING
🚀
Maintainer Growth Kit for TensorRT

Claim this project, add your verified backlink badge to your README, and download milestone cards.

Claim Repo

Star History

Continuous Observations
Interactive star growth chart for pytorch/TensorRT
CSV

Momentum

+75

STARS · LAST 30 DAYS

3

PER DAY

#237

MOST-STARRED Python

Window7 days30 days90 days
Stars gained+21+75+270
Per day333
Forks gained+2+8+21

TensorRT gained 75 stars in the last 30 days, about 3 a day, and now has 3K. It is about 7 years old and has averaged roughly 429 stars a year. It ranks #237 among Python repositories and #1,639 across all languages on GitHubRepo.

Trending Record

TensorRT has maintained a continuous presence across global trending indexes, peaking at #2089. Below is the 30-day activity profile:

💡 Overview

TensorRT is an open-source project written in Python: PyTorch/TorchScript/FX compiler for NVIDIA GPUs using TensorRT.

Engineered for speed, consistency, and developer ease, it solves common hurdles in cuda, deep-learning, jetson. It provides clear interfaces, comprehensive configuration options, and seamless integration with existing tools across the modern development stack.

⚡ Key Features

1

Optimized execution pipeline written in Python for predictable speed.

2

Zero-friction configuration with comprehensive sensible defaults out of the box.

3

Cross-platform runtime support across Linux, macOS, and Windows environments.

4

Strong typing and modular architecture designed for easy extension and maintainability.

5

Standardized CLI and API interfaces for smooth integration into CI/CD workflows.

6

Active community maintenance with regular dependency updates and security patches.

📥 Installation

terminal
$ pip install TensorRT

⚙ System Requirements

Platforms

  • • macOS
  • • Linux
  • • Windows

Runtime & Dependencies

Python >= 3.9, pip, virtualenv

Architecture

x86_64, ARM64 (Apple Silicon & Graviton)

🧠 How It Works

TensorRT coordinates its core functionality through a modular Python pipeline. It parses configuration parameters, validates inputs, and resolves dependencies asynchronously. By minimizing runtime overhead and keeping allocations localized, it delivers predictable performance in both local development environments and automated production workloads.

🎯 Production Use Cases

Production System Integration

Embed TensorRT into Python backend services to handle core application logic.

CI/CD Automated Pipelines

Run automated validation, builds, and integration suites during deployments.

Developer Tooling & Workflows

Accelerate developer onboarding with pre-configured project utilities.

Open Source Extension

Fork and customize internal modules under the repository's open BSD-3-Clause license.

🚀 Getting Started

1

Install TensorRT using your package manager: `pip install TensorRT`

2

Initialize your project workspace or configuration file for TensorRT.

3

Import TensorRT into your codebase or invoke it directly from your terminal.

4

Execute your test suite or run `TensorRT --help` to verify successful setup.

👍 Strengths

Active community backing with 3,001 GitHub stars and verified adoption.
Permissive open-source distribution under the BSD-3-Clause license.
Built in Python for high execution speed and developer familiarity.
Cross-platform compatibility across modern Linux, macOS, and Windows environments.
Clean modular design allowing flexible configuration and pipeline integration.

⚠️ Considerations

Requires familiarity with Python and modern CLI workflows.
Ecosystem extensions may require manual configuration depending on environment constraints.
Active development roadmap means breaking API changes may occur across major versions.

⇄ Alternatives & Direct Competitors

P
public-apis/public-apis ★ 483.8K Python

A collective list of free APIs

Compare ↗
F

:books: Freely available programming books

Compare ↗
P

Curated list of project-based tutorials

Compare ↗
H
NousResearch/hermes-agent ★ 249.8K Python

The agent that grows with you

Compare ↗

👥 Who Should Use This

Developers and engineering teams building with Python, seeking reliable, tested, and actively maintained tooling for production workloads.

🏆 Nearby in the Rankings

pytorch/TensorRT is currently ranked #1,639 by stars across every repository tracked on GitHubRepo. These are adjacent projects:

RankRepositoryLanguageStarsAction
#1,634 androidx/media Java ★ 3K Compare ↗
#1,634 duckdb/ducklake C++ ★ 3K Compare ↗
#1,636 radixark/miles Python ★ 3K Compare ↗
#1,637 headwaymaps/headway Rust ★ 3K Compare ↗
#1,638 nodatime/nodatime C# ★ 3K Compare ↗
#1,639 pytorch/TensorRT This Project Python ★ 3K
#1,640 mickasmt/next-saas-stripe-starter TypeScript ★ 3K Compare ↗
#1,641 rust-lang/hashbrown Rust ★ 3K Compare ↗
#1,642 GoogleCloudPlatform/nodejs-docs-samples JavaScript ★ 3K Compare ↗
#1,643 phcode-dev/phoenix JavaScript ★ 3K Compare ↗
#1,644 elie222/rakazo TypeScript ★ 3K Compare ↗

Frequently Asked Questions

What does TensorRT do? +

PyTorch/TorchScript/FX compiler for NVIDIA GPUs using TensorRT

What language is TensorRT written in? +

The primary language is Python. Topics include: cuda, deep-learning, jetson, libtorch, machine-learning.

Is TensorRT actively maintained? +

Yes, the last recorded push was on Sep 28, 2026 with 342 open issues being tracked.

How many stars does TensorRT have? +

TensorRT has 3,001 stars and 411 forks on GitHub.

How does TensorRT rank among GitHub repositories? +

With 3,001 stars, pytorch/TensorRT is ranked #1,639 globally across all repositories tracked on GitHubRepo and #237 among Python projects.

What license is TensorRT distributed under? +

The repository reports a BSD-3-Clause license. Always verify the repository LICENSE file for legal terms.

From our network
FOR MAINTAINERS

Built something? Put it in front of millions of developers.

We make a short reel about your project and post it across YouTube, Instagram, Threads, and X. Send a link, we do the rest.