Board Archive Editions Search Newsletter

01BoardToday on GitHub
02ArchiveEvery day we kept
03EditionsThe Weekly Star
04SearchThe whole record
05NewsletterEvery Monday

How we pick PrivacyCaptured daily

A field guide to open source.
Follow the work. Keep the record.

How we pick Privacy

A project by Patrick Xin

Board Archive Editions Search Newsletter

01BoardToday on GitHub
02ArchiveEvery day we kept
03EditionsThe Weekly Star
04SearchThe whole record
05NewsletterEvery Monday

How we pick PrivacyCaptured daily

vllm-project/vllm — Starlogs

vllm-project /

vllm

A library for fast LLM inference and serving, offering high-throughput serving and an OpenAI-compatible API server.

library
local-llm
ai-gateway
devops

Stars: 95,444
Sightings: 4
Best rank: No. 3
First seen: Oct 20
Language: Python

Open on GitHub ↗

Each star is a day on a trending board. The higher it ranked that day, the larger and higher the star.

First seen Oct 20, No. 15
Highest at No. 3, Oct 21
Last seen Oct 25, No. 12 · 4 days on the boards

What it's for.

Serve LLMs with high throughput
Run an OpenAI-compatible API server
Scale inference across distributed hardware

How you start.

Install vLLM with uv or pip
Download precompiled kernels for faster startup
Serve models through the OpenAI-compatible API server

Before you begin.

Requires supported hardware such as NVIDIA, AMD or Intel GPUs, or x86/ARM/PowerPC CPUs.
Other accelerators such as TPUs, Gaudi or Ascend rely on separate hardware plugins.

Stars+1,137 since first seen

counted on a capture dayjoined straight between counts

Sightings4 days on the boards

no boards keptnot on a boardfewer starsmore

Nearby

Close by description, tags or language

NVIDIA-NeMo
Switchyard
Routes each LLM call to the cheapest capable model via gateway plugins, an embeddable library, or a standalone proxy.

MiaAI-Lab

DeepSeek-v4-Flash-DSpark-2x-DGX-Spark

Deployment recipe for serving DeepSeek-V4-Flash-Vision-Exp with vLLM across two DGX Spark nodes with 1M-token context and image input.

BerriAI

litellm

Open source AI gateway and Python SDK that provides a unified OpenAI-format interface to call 100+ LLM providers.