Package search

vllm

Summary: A high-throughput and memory-efficient inference and serving engine for LLMs
Author: vLLM Team
Homepage: https://github.com/vllm-project/vllm
Source: https://github.com/vllm-project/vllm
Repo profile
Number of releases: 58
First release: 0.0.1 on 2023-06-19T08:21:42
Latest release: 0.8.5.post1 on 2025-05-02T22:31:02

Dependencies of vllm (62).
Dependency	Optional
aiohttp	false
blake3	false
cachetools	false
cloudpickle	false
compressed-tensors	false
depyf	false
einops	false
fastapi	false
filelock	false
gguf	false
huggingface-hub	false
importlib_metadata	false
lark	false
llguidance	false
lm-format-enforcer	false
mistral_common	false
msgspec	false
ninja	false
numba	false
numpy	false
openai	false
opencv-python-headless	false
opentelemetry-api	false
opentelemetry-exporter-otlp	false
opentelemetry-sdk	false
opentelemetry-semantic-conventions-ai	false
outlines	false
partial-json-parser	false
pillow	false
prometheus_client	false
prometheus-fastapi-instrumentator	false
protobuf	false
psutil	false
py-cpuinfo	false
pydantic	false
python-json-logger	false
pyyaml	false
pyzmq	false
ray	false
requests	false
scipy	false
sentencepiece	false
setuptools	false
six	false
tiktoken	false
tokenizers	false
torch	false
torchaudio	false
torchvision	false
tqdm	false
transformers	false
typing_extensions	false
watchfiles	false
xformers	false
xgrammar	false
boto3	true
fastsafetensors	true
librosa	true
runai-model-streamer	true
runai-model-streamer-s3	true
soundfile	true
tensorizer	true