What is ART built with?

ART is a Python repo built with FastAPI, AIOHTTP, and weave. It uses AI through apache-tvm-ffi, causal-conv1d, flash-attn-4, flash-linear-attention, flashinfer-python, and 12 more and validates its data with Pydantic and msgspec. It has 345,462 written lines, with 391 test files.

Read on November 13, 2026, at commit 534620a. Read it as markdown

The card says what it's built with. Its whole map opens it up: every part of the code and how it connects to the rest.

Backend

The base 1

AI 15

3 more
  • ml-dtypes
  • quack-kernels
  • torchtune

Services 2

Support

The base 2

Validation 2

Tools 7

  • Typer5 files
  • ninja
  • pybind11
  • Setuptools
  • uv
  • typing-extensions13 files
  • tilelang

Other 11

  • NumPy15 files
  • Polars8 files
  • SciPy1 file
  • nvidia-ml-py
  • nvidia-resiliency-ext
  • python-dotenv23 files
  • nest-asyncio2 files
  • orjson1 file
  • requests1 file
  • setproctitle1 file
  • tblib1 file

Other repos with a similar stack

Ordered by how much of its stack they share with ART, then by stars.

Frequently asked questions

What language is ART written in?
ART is written in Python, with 345,462 written lines.
What framework does ART use?
ART is built with FastAPI, AIOHTTP, and weave.
Does ART use AI?
Yes. ART uses AI through apache-tvm-ffi, causal-conv1d, flash-attn-4, flash-linear-attention, flashinfer-python, megatron-core, ml-dtypes, NVIDIA, quack-kernels, torchtune, unsloth-zoo, torch, transformers, litellm, peft, OpenAI, and Anthropic.
Does ART have tests?
Yes. ART has 391 test files.
Which repos have a stack like ART's?
The ones that share most of its stack:sgl-project/sglangvllm-project/vllmbojieli/ai-agent-book

See inside your code, today.

Paste your repo and see it drawn, part by part: a medium one in under a minute.

Free, no account.

Or look inside SeeRepo's own →

Is it your repo? Add its badge, keep it up to date or take it down →