Footer

Your go-to destination for discovering trending open-source projects and uncovering the insights that matter

Legal

  • Terms and conditions
  • Privacy policy

More

  • Pricing
  • Advertise
  • onedump
  • gosnakego
  • Contact us

Social

  • X (Twitter)
  • GitHub
Trendshift owl
What they build speaks louder than words
© 2026 Trendshift, created by Julian Li with ❤️
SignalAdvertise
Trendshift
SignalAdvertise
Sign in
Trendshift owl

Discover trends that matter

SignalAdvertise
DailyWeeklyMonthlyYearly
Live mentionsTopics
RepositoriesDevelopers
InsightsStats

Omnigraph
Omnigraph
OfinTech Desk
OfinTech Desk
Busbar
Busbar
Bifrost
Bifrost
It's a plan
It's a plan

vllm-project/vllm

A high-throughput and memory-efficient inference and serving engine for LLMs

#LLM inference
Visit GitHub
Python
97.3k
25k
2,758 contributors
Apache License 2.0
website
AchievementsView all →
FeaturedAdvertise to 180K+ monthly visitors
Omnigraph

Versioned graph database on object storage. Git-style workflows and multimodal retrieval

OfinTech Desk

An open-source AI workspace for finance and accounting teams, in which every figure is traced to its source

Busbar

Control what AI can do before it acts

Bifrost

Fastest Enterprise AI Gateway for Routing, Governing, and Securing all your AI Traffic

It's a plan

Open-source Linear alternative

An issue tracker where AI agents take the work you assign them

Momentum

How fast vllm-project/vllm is gaining stars, and where it lands on Trendshift and GitHub Trending

Activity and trending

Stars, forks and merged pull requests each day or month, lined up with the repository's rank on Trendshift Trending

Sep 22 – Nov 21, 2026

Stars +4,936 in 61 days

0200400600
424

Forks +2,897 in 61 days

050100
73

Merged PRs 4,284 in 61 days

050100150
126

Trendshift Trendingwhere the repository ranked each day it made the top 25

Top 25 in Python

1913
Sep 22Sep 30Oct 8Oct 16Oct 24Nov 1Nov 9Nov 17
The left and right arrow keys step from day to day; Home and End go to the first and the last

Stars, forks, merged pull requests and Trendshift Trending rank, by day
DayStars gainedForks gainedMerged PRsPython
Tue, Sep 22, 2026+85+4563Not trending
Wed, Sep 23, 2026+95+6055Not trending
Thu, Sep 24, 2026+97+4972Not trending
Fri, Sep 25, 2026+43+4177Not trending
Sat, Sep 26, 2026+43+2925Not trending
Sun, Sep 27, 2026+76+3019Not trending
Mon, Sep 28, 2026+83+5665Not trending
Tue, Sep 29, 2026+74+4185Not trending
Wed, Sep 30, 2026+47+5391Not trending
Thu, Oct 1, 2026+33+5368Not trending
Fri, Oct 2, 2026+42+2969Not trending
Sat, Oct 3, 2026+46+2418Not trending
Sun, Oct 4, 2026+51+3618Not trending
Mon, Oct 5, 2026+46+4050Not trending
Tue, Oct 6, 2026+70+5069Not trending
Wed, Oct 7, 2026+51+3858Not trending
Thu, Oct 8, 2026+64+5491Not trending
Fri, Oct 9, 2026+52+4976Not trending
Sat, Oct 10, 2026+63+4634Not trending
Sun, Oct 11, 2026+54+3821Not trending
Mon, Oct 12, 2026+86+4075Not trending
Tue, Oct 13, 2026+69+5582Not trending
Wed, Oct 14, 2026+66+4888Not trending
Thu, Oct 15, 2026+70+4289Not trending
Fri, Oct 16, 2026+76+3449Not trending
Sat, Oct 17, 2026+61+5448Not trending
Sun, Oct 18, 2026+70+4722Not trending
Mon, Oct 19, 2026+82+6283Not trending
Tue, Oct 20, 2026+264+7086#19
Wed, Oct 21, 2026+424+7269#13
Thu, Oct 22, 2026+303+7389Not trending
Fri, Oct 23, 2026+98+5991Not trending
Sat, Oct 24, 2026+93+4326Not trending
Sun, Oct 25, 2026+145+4934Not trending
Mon, Oct 26, 2026+136+67100Not trending
Tue, Oct 27, 2026+88+66103Not trending
Wed, Oct 28, 2026+77+4877Not trending
Thu, Oct 29, 2026+79+41100Not trending
Fri, Oct 30, 2026+69+5497Not trending
Sat, Oct 31, 2026+68+3040Not trending
Sun, Nov 1, 2026+75+3910Not trending
Mon, Nov 2, 2026+86+6190Not trending
Tue, Nov 3, 2026+75+55126Not trending
Wed, Nov 4, 2026+89+62111Not trending
Thu, Nov 5, 2026+85+60100Not trending
Fri, Nov 6, 2026+64+39105Not trending
Sat, Nov 7, 2026+52+4335Not trending
Sun, Nov 8, 2026+31+3035Not trending
Mon, Nov 9, 2026+64+5699Not trending
Tue, Nov 10, 2026+66+5377Not trending
Wed, Nov 11, 2026+82+4987Not trending
Thu, Nov 12, 2026+63+4994Not trending
Fri, Nov 13, 2026+40+48103Not trending
Sat, Nov 14, 2026+63+4749Not trending
Sun, Nov 15, 2026+36+3526Not trending
Mon, Nov 16, 2026+78+50113Not trending
Tue, Nov 17, 2026+63+4391Not trending
Wed, Nov 18, 2026+69+53111Not trending
Thu, Nov 19, 2026+38+58124Not trending
Fri, Nov 20, 2026+61+49122Not trending
Sat, Nov 21, 2026+17+34Not trending
Total+4,936+2,8974,284
  • 1#1 on the list
  • 7Rank that day
  • Not trending
1
  • GitHub trending history

    Where the repository ranked on GitHub Trending, one dot for each day it was on a list

    Sep 16, 2023 – Oct 21, 2026Best #2

    #1#5#10#15#20#25
    Top 10
    Oct 2023AprOctAprOctAprOct
    The left and right arrow keys step from day to day; Home and End go to the first and the last

    Mentions and achievements

    Where people are talking about vllm-project/vllm, and every badge it has earned
    1. Nov 202 posts
      • My first PR to vLLM... I'd say, pretty much just practice, but here it is:

        Read full on X, posted by @pupeno

        @pupeno · x.com

      • 100K-token prompts spike vLLM tail latency even when the GPU isn't slow. The output loop was re-decoding the entire prompt on token 1. Walking back 5 text tokens cut 100K detokenization 41.6 -> 0.007…

        Read full on X, posted by @GarryLing1

        @GarryLing1 · x.com

    2. Nov 192 posts
      • GLM-5.3-Flash DeepGEMM MegaMoE support for Hopper just landed with good speedups on AgentX and native FP8 weights support in vLLM! Give it a try!

        Read full on X, posted by @Simon_Vt

        @Simon_Vt · x.com

      • Enabling cudagraph_mm_encoder=True on Kimi-VL in vLLM silently corrupts OCR ("J" -> "j") with no errors. MoonViT runs on pre-merge patches (4x tokens), yet capture baked…

        Read full on X, posted by @GarryLing1

        @GarryLing1 · x.com

    3. Nov 181 post5 posts
      • vLLM 修复了一种 LoRA 混合批次错误:启用 CUDA graphs 和 specialize_active_lora 时,LoRA 请求可能得到基础模型输出。修复已于 11 月 18 日合并至主分支。

        Read full on X, posted by @CarlWang0524

        @CarlWang0524 · x.com

      • Running an LLM locally is easy. Serving it to real users is the hard part. vLLM (~97k stars, Apache-2.0) uses PagedAttention + continuous batching and ships an OpenAI-compatible API. Local dev:…

        Read full on X, posted by @touseefdevv

        @touseefdevv · x.com

      • Hy4 flash in NVFP4, on day 0. Tencent's 220B open-weight model with 12B active now runs on two DGX Sparks. The 4-bit build is 165 GB. vLLM splits it across both Sparks: 36-38 tok/s on code and JSON,…

        Read full on X, posted by @atasrb

        @atasrb · x.com

      • 8/9 BF16 Loop weights: ~17.6 GB; 4-bit Qwen: ~15 GB. Same memory class, and Loop runs 56 layer passes/token. “9B” buys neither a smaller footprint nor 9B-class speed here. ~26 tok/s BF16 decode; no…

        Read full on X, posted by @lasse_ai

        @lasse_ai · x.com

      • vLLM prefix caching can OOM mid-serving from one line: blocks[dst] = blocks[src]. PyTorch materializes a GiB-scale gather temp never reserved at warmup. PR #61051 swaps in a zero-temp Triton kernel…

        Read full on X, posted by @GarryLing1

        @GarryLing1 · x.com

    4. Nov 171 post
      • In Zhihu contributor Jinzhen’s 8192 × 8192 × 8192 dense matrix-multiplication benchmark, a revised Humming MXFP4 × FP8 kernel reached 1,293.64 TFLOPS on H200, up from 681.80, about 1.90× the previous…

        Read full on X, posted by @ZhihuFrontier

        @ZhihuFrontier · x.com

    You might also like

    Repositories close to vllm-project/vllm
    FeaturedAdvertise to 180K+ monthly visitors
    OfinTech Desk

    An open-source AI workspace for finance and accounting teams, in which every figure is traced to its source

    Omnigraph

    Versioned graph database on object storage. Git-style workflows and multimodal retrieval

    It's a plan

    Open-source Linear alternative

    An issue tracker where AI agents take the work you assign them

    Bifrost

    Fastest Enterprise AI Gateway for Routing, Governing, and Securing all your AI Traffic

    Busbar

    Control what AI can do before it acts

    local-inference-lab/vllm

    Python

    A high-throughput and memory-efficient inference and serving engine for LLMs

    4749

    inferflow/inferflow

    C++

    Inferflow is an efficient and highly configurable inference engine for large language models (LLMs).

    25324

    hpcaitech/SwiftInfer

    Python

    Efficient AI Inference & Serving

    47031

    harleyszhang/rapid_llm

    Python

    RapidLLM is an LLM inference framework with continuous batching, tensor and data parallelism, quantized weights, and with pluggable Triton/CUDA kernels.

    19833

    gty111/gLLM

    Python

    An Efficient and Versatile Inference Engine for Distributed LLM Serving

    7410

    scaleapi/llm-engine

    Python

    Scale LLM Engine public repository

    841100

    mlc-ai/mlc-llm

    Python

    Universal LLM Deployment Engine with ML Compilation

    23.4k2.4k

    InternLM/lmdeploy

    Python

    LMDeploy is a toolkit for compressing, deploying, and serving LLMs.

    8.2k841

    EricLBuehler/candle-vllm

    Rust

    Efficent platform for inference and serving local LLMs including an OpenAI compatible API server.

    74293