Source: https://seerepo.com/compare/NVIDIA/TensorRT-LLM/vs/sgl-project/sglang
Markdown source: https://seerepo.com/compare/NVIDIA/TensorRT-LLM/vs/sgl-project/sglang.md
Description: TensorRT-LLM and sglang share 29 technologies: FastAPI, blobfile, pyzmq, OpenAI, uvicorn, apache-tvm-ffi, cuda-python, flashinfer-python, NVIDIA, openai-harmony, quack-kernels, sentencepiece, tiktoken, torch-c-dlpack-ext, torch, transformers, einops, torchvision, datasets, xgrammar, mistral-common, partial-json-parser, llguidance, pillow, Pydantic, msgspec, psutil, prometheus-client, and OpenTelemetry. Both are written in Python.

# TensorRT-LLM vs sglang

TensorRT-LLM and sglang share 29 technologies: FastAPI, blobfile, pyzmq, OpenAI, uvicorn, apache-tvm-ffi, cuda-python, flashinfer-python, NVIDIA, openai-harmony, quack-kernels, sentencepiece, tiktoken, torch-c-dlpack-ext, torch, transformers, einops, torchvision, datasets, xgrammar, mistral-common, partial-json-parser, llguidance, pillow, Pydantic, msgspec, psutil, prometheus-client, and OpenTelemetry. Both are written in Python.

- [NVIDIA/TensorRT-LLM](https://seerepo.com/r/NVIDIA/TensorRT-LLM): 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.
- [sgl-project/sglang](https://seerepo.com/r/sgl-project/sglang): SGLang is a high-performance serving framework for large language models and multimodal models.

## What they share and what sets them apart

- In common: FastAPI, blobfile, pyzmq, OpenAI, uvicorn, apache-tvm-ffi, cuda-python, flashinfer-python, NVIDIA, openai-harmony, quack-kernels, sentencepiece, tiktoken, torch-c-dlpack-ext, torch, transformers, einops, torchvision, datasets, xgrammar, mistral-common, partial-json-parser, llguidance, pillow, Pydantic, msgspec, psutil, prometheus-client, OpenTelemetry
- Only in TensorRT-LLM: h5py, anyio, starlette, Rich, matplotlib, Plotly, accelerate, flash-linear-attention, nvidia-cudnn-frontend, openai-codex, optimum, rouge, rouge-score, torchao, triton, diffusers, safetensors, flash-attn, librosa, MCP, tritonclient, evaluate, lm-eval, peft, jsonlines, jinja2, mako, jsonschema, opentelemetry-api, opentelemetry-semantic-conventions-ai, boto3, opentelemetry-sdk, prometheus-fastapi-instrumentator
- Only in sglang: axum, tower, axum-server, Redis, Anthropic, Tokio, AIOHTTP, wasmtime, indicatif, jsonwebtoken, humming-kernels, compressed-tensors, gguf, tokenizers, ModelScope, tch, torchaudio, hf-hub, rmcp, timm, minijinja, av, torchcodec, image, validator, tracing, tracing-subscriber, k8s-openapi, kube, metrics, tracing-appender, metrics-exporter-prometheus, opentelemetry-otlp, tracing-log, tracing-opentelemetry

## Side by side

| | TensorRT-LLM | sglang |
|---|---|---|
| Size | 2,884,778 lines written, 3,120 code files | 3,312,331 lines written, 4,627 code files |
| Who writes it | 1.7% of its commits signed by AI. With Claude Code, CodeRabbit, GitHub Copilot, and Cursor | 8.7% of its commits signed by AI. With Claude Code, Gemini CLI, Cursor, GitHub Copilot, Devin, Codex, CodeRabbit, and OpenHands |
| Activity | Last push November 8, 2026, 10,997 commits, 15,038 stars | Last push November 6, 2026, 21,830 commits, 38,542 stars |
| Tests | 2,703 test files. 58.6% of its code files have a test that imports them | 4,212 test files. 59.6% of its code files have a test that imports them |

- Size: Lines written by people and agents, without generated ones; code files, without tests.
- Who writes it: A commit counts as AI's when it carries its signature: a Co-Authored-By line, its author name or the merge of its branch.
- Activity: From GitHub and its history, on the day SeeRepo read it.
- Tests: The files that test the code, and the share of code files a test imports.

## Frequently asked questions

### Which is bigger, TensorRT-LLM or sglang?

sglang, with 3,312,331 lines written; TensorRT-LLM has 2,884,778.

The proof: [sgl-project/sglang's code at commit 0afb736](https://github.com/sgl-project/sglang/tree/0afb736d0aab96e3ab80b4c54a376213a1db67a6)

### Which has more GitHub stars, TensorRT-LLM or sglang?

sglang, with 38,542 stars; TensorRT-LLM has 15,038.

The proof: [sgl-project/sglang on GitHub](https://github.com/sgl-project/sglang)

### Which has more tests, TensorRT-LLM or sglang?

sglang, with 4,212 test files; TensorRT-LLM has 2,703.

The proof: [sgl-project/sglang's code at commit 0afb736](https://github.com/sgl-project/sglang/tree/0afb736d0aab96e3ab80b4c54a376213a1db67a6)

### Which was updated more recently, TensorRT-LLM or sglang?

TensorRT-LLM: its last push was on November 8, 2026, and sglang's on November 6, 2026.

The proof: [NVIDIA/TensorRT-LLM's commits on GitHub](https://github.com/NVIDIA/TensorRT-LLM/commits)

### Which has more commits signed by AI, TensorRT-LLM or sglang?

sglang: an AI signed 8.7% of its commits, with Claude Code, Gemini CLI, Cursor, GitHub Copilot, Devin, Codex, CodeRabbit, and OpenHands; in TensorRT-LLM, 1.7%.

The proof: [the commits each tool signed in sgl-project/sglang](https://seerepo.com/r/sgl-project/sglang/ai)

## Sources

- NVIDIA/TensorRT-LLM: Read on November 8, 2026, at commit b433473 (https://github.com/NVIDIA/TensorRT-LLM/tree/b433473ba0d0d49764f4d2fb3d45c82716b946dc)
- sgl-project/sglang: Read on November 6, 2026, at commit 0afb736 (https://github.com/sgl-project/sglang/tree/0afb736d0aab96e3ab80b4c54a376213a1db67a6)

## 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.