Source: https://seerepo.com/compare/InternLM/lmdeploy/vs/vllm-project/vllm
Markdown source: https://seerepo.com/compare/InternLM/lmdeploy/vs/vllm-project/vllm.md
Description: lmdeploy and vllm share 22 technologies: FastAPI, pyzmq, OpenAI, Ray, AIOHTTP, apache-tvm-ffi, openai-harmony, partial-json-parser, torch, transformers, triton, xgrammar, safetensors, torchvision, einops, sentencepiece, tiktoken, timm, markdown, pillow, Pydantic, and prometheus-client. Both are written in Python.

# lmdeploy vs vllm

lmdeploy and vllm share 22 technologies: FastAPI, pyzmq, OpenAI, Ray, AIOHTTP, apache-tvm-ffi, openai-harmony, partial-json-parser, torch, transformers, triton, xgrammar, safetensors, torchvision, einops, sentencepiece, tiktoken, timm, markdown, pillow, Pydantic, and prometheus-client. Both are written in Python.

- [InternLM/lmdeploy](https://seerepo.com/r/InternLM/lmdeploy): LMDeploy is a toolkit for compressing, deploying, and serving LLMs.
- [vllm-project/vllm](https://seerepo.com/r/vllm-project/vllm): A high-throughput and memory-efficient inference and serving engine for LLMs

## What they share and what sets them apart

- In common: FastAPI, pyzmq, OpenAI, Ray, AIOHTTP, apache-tvm-ffi, openai-harmony, partial-json-parser, torch, transformers, triton, xgrammar, safetensors, torchvision, einops, sentencepiece, tiktoken, timm, markdown, pillow, Pydantic, prometheus-client
- Only in lmdeploy: Anthropic, uvicorn, sphinx-book-theme, matplotlib, flash-linear-attention, qwen-vl-utils, rdkit, transformers-stream-generator, accelerate, peft, torch-npu, datasets, myst-parser
- Only in vllm: axum, gitdb, Hugging Face, Tokio, starlette, tower, mkdocs-material, indicatif, AMD, auto-round-lib, conch-triton-kernels, cuda-bindings, cuda-python, cupy-cuda12x, cupy-cuda13x, flashinfer-python, humming-kernels, lm-format-enforcer, mooncake-transfer-engine, mooncake-transfer-engine-rocm, NVIDIA, nvidia-cudnn-frontend, oss-harmony, quack-kernels, vllm-bench, vllm-chat, vllm-engine-core-client, vllm-llm, vllm-managed-engine, vllm-parser, vllm-proto, vllm-server, vllm-text, vllm-tokenizer, xgrammar-structural-tag, llm-multimodal, compressed-tensors, mistral-common, lmcache, torchaudio, vllm-xpu-kernels, runai-model-streamer, tokenizers, tokenspeed-mla, llguidance, MCP, outlines-core, tpu-inference, fastokens, fastsafetensors, hf-hub, instanttensor, nixl, riptoken, tekken, tensorizer, tiktoken-rs, markupsafe, mkdocs-material-extensions, parquet, properdocs, pymdown-extensions, selectolax, xlsxwriter, image, jinja2, minijinja, torchcodec, minijinja-contrib, annotated-types, msgspec, validator, opentelemetry-semantic-conventions-ai, python-json-logger, vllm-metrics, vllm-tracing, tracing, psutil, tracing-futures, tracing-subscriber, model-hosting-container-standards, OpenTelemetry, prometheus-fastapi-instrumentator, tracing-log

## Side by side

| | lmdeploy | vllm |
|---|---|---|
| Size | 299,565 lines written, 789 code files | 2,606,460 lines written, 3,630 code files |
| Who writes it | 1.2% of its commits signed by AI. With Claude Code, GitHub Copilot, Cursor, and OpenHands | 14.7% of its commits signed by AI. With Claude Code, Codex, Gemini CLI, Cursor, GitHub Copilot, Devin, Mistral Vibe, CodeRabbit, Factory, and Qwen Code |
| Activity | Last push November 17, 2026, 2,189 commits, 8,149 stars | Last push November 17, 2026, 25,558 commits, 97,060 stars |
| Tests | 336 test files. 44.1% of its code files have a test that imports them | 2,586 test files. 60.7% 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, lmdeploy or vllm?

vllm, with 2,606,460 lines written; lmdeploy has 299,565.

The proof: [vllm-project/vllm's code at commit 61c3895](https://github.com/vllm-project/vllm/tree/61c3895fc937b4ffc2a59abe188959de1e459c8a)

### Which has more GitHub stars, lmdeploy or vllm?

vllm, with 97,060 stars; lmdeploy has 8,149.

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

### Which has more tests, lmdeploy or vllm?

vllm, with 2,586 test files; lmdeploy has 336.

The proof: [vllm-project/vllm's code at commit 61c3895](https://github.com/vllm-project/vllm/tree/61c3895fc937b4ffc2a59abe188959de1e459c8a)

### Which was updated more recently, lmdeploy or vllm?

Both got their last push on November 17, 2026.

The proof: [InternLM/lmdeploy's commits on GitHub](https://github.com/InternLM/lmdeploy/commits)

### Which has more commits signed by AI, lmdeploy or vllm?

vllm: an AI signed 14.7% of its commits, with Claude Code, Codex, Gemini CLI, Cursor, GitHub Copilot, Devin, Mistral Vibe, CodeRabbit, Factory, and Qwen Code; in lmdeploy, 1.2%.

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

## Sources

- InternLM/lmdeploy: Read on November 17, 2026, at commit fc33f69 (https://github.com/InternLM/lmdeploy/tree/fc33f69736ff48c10ad720593b021dcdaedcb80d)
- vllm-project/vllm: Read on November 17, 2026, at commit 61c3895 (https://github.com/vllm-project/vllm/tree/61c3895fc937b4ffc2a59abe188959de1e459c8a)

## 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.
