Source: https://seerepo.com/r/NVIDIA/TensorRT-LLM
Markdown source: https://seerepo.com/r/NVIDIA/TensorRT-LLM.md
Description: NVIDIA/TensorRT-LLM is a Python repo with 2,884,778 lines written. TensorRT-LLM is built with FastAPI, anyio, uvicorn, and starlette and uses OpenAI.

# What's inside NVIDIA/TensorRT-LLM

NVIDIA/TensorRT-LLM is a Python repo with 2,884,778 lines written. TensorRT-LLM is built with FastAPI, anyio, uvicorn, and starlette and uses OpenAI. AI signed 1.7% of the 10,997 commits of TensorRT-LLM.

Its description on GitHub: “TensorRT LLM provides users with an easy-to-use Python API to define Large Language Models (LLMs) and supports state-of-the-art optimizations to perform inference efficiently on NVIDIA GPUs. TensorRT LLM also contains components to create Python and C++ runtimes that orchestrate the inference execution in a performant way.”

Python · 15,038 stars · GitHub: https://github.com/NVIDIA/TensorRT-LLM

Read on November 8, 2026, at commit b433473.

## More on this repo

- [What is TensorRT-LLM built with?](https://seerepo.com/r/NVIDIA/TensorRT-LLM/built): TensorRT-LLM is built with FastAPI, anyio, uvicorn, and starlette and uses OpenAI.
- [How much of TensorRT-LLM did AI write?](https://seerepo.com/r/NVIDIA/TensorRT-LLM/ai): AI signed 1.7% of the 10,997 commits of TensorRT-LLM.

## In figures

- 2,884,778 lines written: The ones people and agents wrote: product, tests, schemas and tooling. Generated code doesn’t count.
- 3,120 source files: The same as the lines: product, tests, schemas and tooling. What’s generated doesn’t count.
- 2,703 test files: The files that test the code.
- 62 technologies: Each service with the file and line that prove it.

SeeRepo read the facts in its code and kept none of it. AI only helps sort its packages and folders by their names, and never sees its code.

## Frequently asked questions

### What is NVIDIA/TensorRT-LLM?

NVIDIA/TensorRT-LLM is a Python repo with 2,884,778 lines written. Its description on GitHub: “TensorRT LLM provides users with an easy-to-use Python API to define Large Language Models (LLMs) and supports state-of-the-art optimizations to perform inference efficiently on NVIDIA GPUs. TensorRT LLM also contains components to create Python and C++ runtimes that orchestrate the inference execution in a performant way.”

The proof: [its page on GitHub](https://github.com/NVIDIA/TensorRT-LLM)

### What is TensorRT-LLM built with?

TensorRT-LLM is built with FastAPI, anyio, uvicorn, and starlette and uses OpenAI.

The proof: [each service with its file and line](https://seerepo.com/r/NVIDIA/TensorRT-LLM/built)

### How much of TensorRT-LLM did AI write?

AI signed 1.7% of the 10,997 commits of TensorRT-LLM.

The proof: [the commits each tool signed](https://seerepo.com/r/NVIDIA/TensorRT-LLM/ai)

### How many stars does TensorRT-LLM have on GitHub?

TensorRT-LLM had 15,038 stars on GitHub on November 8, 2026.

The proof: [its page on GitHub](https://github.com/NVIDIA/TensorRT-LLM)

## Add this badge to your README

It links to this page and updates itself each time SeeRepo reads the repo.

```markdown
[![NVIDIA/TensorRT-LLM on SeeRepo](https://seerepo.com/badge/NVIDIA/TensorRT-LLM.svg)](https://seerepo.com/r/NVIDIA/TensorRT-LLM?de=insignia)
```

## What is your repo built with?

Any public repo, with no account: https://seerepo.com/r/owner/repo/built, with the owner and name it has on GitHub.