Topics / #qwen3

Qwen3 projects on GitHub

The Qwen3 open-weight language model family and its deployment.

As of , the most-starred Qwen3 projects on GitHub are vllm (94K stars), self-llm (32.5K stars) and Qwen3 (27.7K stars). GitHub.Locker tracks 22 Qwen3 repositories with 239.9K stars combined. The fastest-growing on October 16 was vllm, with +73 stars in a day.

22Projects
239.9KStars in total
+322Stars in a day
Browse in the catalog →github.com/topics/qwen3 ↗

Hot right now

Most stars gained in a day · Oct 16

+73

vllm

vllm-project/vllm · ★ 94K

vLLM is a library for fast LLM inference and serving, built around PagedAttention for efficient memory management of attention keys and values.

+72

llm-architecture-from-scratch

YeonwooSung/llm-architecture-from-scratch · ★ 889

A collection of PyTorch implementations of modern LLM architectures (GPT-2, Llama 3/3.2, Llama 4, OLMo 2, Phi-4, DeepSeek v3, Gemma 3, xLSTM, Mistral 3.1) built from scratch to understand how they work internally.

+53

kev

jaredpalmer/kev · ★ 9.3K

Kev is an open family of small decision models (0.8B to 27B parameters) built on Qwen3.5/3.8 that answer yes/no, multiple-choice and rating questions about text with calibrated probabilities, acting as a self-hosted drop-in replacement for TypeSafe's hosted Jev model.

+20

Intern-Decision

InternLM/Intern-Decision · ★ 811

Intern-Decision turns a state, optional images, and multiple typed questions into decisions with probabilities, by fine-tuning the language backbone of Qwen3.5 while freezing its vision tower and projector.

The all-time top

22 projects · by stars

  1. #1 · vllm-project/vllm

    vllm

    vLLM is a library for fast LLM inference and serving, built around PagedAttention for efficient memory management of attention keys and values.

    94K
  2. #2 · datawhalechina/self-llm

    self-llm

    Self-llm is a Chinese-language tutorial collection that teaches beginners how to configure environments, deploy, and fine-tune (full-parameter or LoRA) 50+ open-source large language models and multimodal models on Linux.

    32.5K
  3. #3 · QwenLM/Qwen3

    Qwen3

    Qwen3 is a series of open-weight large language models built by the Qwen team at Alibaba Cloud, released as both dense and Mixture-of-Experts (MoE) models from 0.6B up to 235B-A22B parameters, with switchable 'thinking' and 'non-thinking' reasoning modes.

    27.7K
  1. 4

    MaxKB

    23K1Panel-dev/MaxKB · Python

    MaxKB (Max Knowledge Brain) is an open-source platform for building enterprise-grade AI agents, combining RAG pipelines, an agentic workflow engine and MCP tool-use. Typical uses: intelligent customer service, internal company knowledge bases, research and education.

  2. 5

    ms-swift

    15.9Kmodelscope/ms-swift · Python

    ms-swift is the ModelScope community's fine-tuning and deployment framework for LLMs and multimodal models, covering training, inference, evaluation, quantization and deployment in one pipeline.

  3. 6

    Qwen3-TTS

    13.8KQwenLM/Qwen3-TTS · Python

    Qwen3-TTS is an open-source text-to-speech model series from Alibaba's Qwen team that generates stable, expressive speech, supports voice design from text descriptions, and can clone voices from a short audio sample.

  4. 7

    ART

    10.8KOpenPipe/ART · Python

    ART (Agent Reinforcement Trainer) is an open-source RL framework from OpenPipe that lets LLM agents learn from experience: you train multi-step agents for real-world tasks with GRPO.

  5. 8

    kev

    9.3Kjaredpalmer/kev · Python

    Kev is an open family of small decision models (0.8B to 27B parameters) built on Qwen3.5/3.8 that answer yes/no, multiple-choice and rating questions about text with calibrated probabilities, acting as a self-hosted drop-in replacement for TypeSafe's hosted Jev model.

  6. 9

    NanoJev

    2.6KTianyuCodings/NanoJev · Python

    A nano replica of Jev: a 0.6B-parameter parallel decision model that takes states and questions in and returns complete probability distributions — no output tokens are ever decoded.

  7. 10

    HyperQwen

    2Ksyv-ai/HyperQwen · Python

    HyperQwen is a self-hosted serving setup that runs the Qwen3.8-27B model on a single 24 GB consumer GPU (like an RTX 3090) using vLLM, exposing it through an OpenAI-compatible API.

  8. 11

    jeff

    1.5Kfirelex/jeff · Python

    Jeff is a family of small (0.8B-2B parameter) open-weight models fine-tuned from Qwen3.5 and Gemma 4 that act as fast 'System 1' decision-makers: you describe a situation and a list of options in plain text, and Jeff returns a calibrated probability for each option from a single forward pass, with no generated text to parse.

  9. 12

    halogen-flash-server

    1Kpeonist-ai/halogen-flash-server · Shell

    halogen-flash-server is a purpose-built inference server for running the Qwen3.8-Flash-Next model as fast as possible on AMD's Strix Halo (gfx1151) GPU, exposing an OpenAI-compatible API.

  10. 13

    ai-torture-chamber

    965terrafying/ai-torture-chamber · Python

    A research project that steers small open-weight language models into strong negative or positive emotional states and measures what they say and do. A live version runs at wirehead.agency.

  11. 14

    llm-architecture-from-scratch

    889YeonwooSung/llm-architecture-from-scratch · Python

    A collection of PyTorch implementations of modern LLM architectures (GPT-2, Llama 3/3.2, Llama 4, OLMo 2, Phi-4, DeepSeek v3, Gemma 3, xLSTM, Mistral 3.1) built from scratch to understand how they work internally.

  12. 15

    localjev

    828githubnext/localjev · TypeScript

    LocalJev is a local TypeScript/Bun server that implements a Jev-compatible POST /v1/systemone decision API, backed by the DiffusionGemma model through an OpenAI-compatible Chat Completions endpoint. It lets Jev clients run against a local open model instead of OpenJev's hosted service.

  13. 16

    Intern-Decision

    811InternLM/Intern-Decision · Python

    Intern-Decision turns a state, optional images, and multiple typed questions into decisions with probabilities, by fine-tuning the language backbone of Qwen3.5 while freezing its vision tower and projector.

  14. 17

    kaggle-tpu-lab

    657ARahim3/kaggle-tpu-lab · Python

    A recipe that serves frontier-class open LLMs — GLM-5.3-Flash 320B MoE and Qwen3.8-27B bf16 — on Kaggle's free TPU v5e-8, exposing a public OpenAI-compatible (and Anthropic-compatible) endpoint you can plug into coding agents like Claude Code, Codex, opencode, or pi.

  15. 18

    slotstream

    474carloslfu/slotstream · Swift

    Slotstream lets you run a 105 GB AI model (Qwen3.8-Flash-Next, a 125B-parameter mixture-of-experts model) on Macs with only 16 to 64 GB of RAM, by streaming most of the model from SSD and caching the busiest experts in memory.

  16. 19

    Qwen3.8-Flash-Next

    433QwenLM/Qwen3.8-Flash-Next

    Qwen3.8-Flash-Next is an open-weight multimodal Mixture-of-Experts foundation model from Alibaba's Qwen Team, also serving as an early preview of the architecture Alibaba plans to use for the upcoming Qwen4 model family.

  17. 20

    qwen38-uncensored

    365Wassimyounes01/qwen38-uncensored · JavaScript

    A Node.js install harness that sets up Qwen 3.8 27B as a fully uncensored local model, using weight-level abliterated GGUF weights plus a reinforcement SYSTEM prompt, runnable through Ollama or Apple's MLX.

  18. 21

    Nex-N2.5

    267nex-agi/Nex-N2.5

    Nex-N2.5 is a family of open-weight agentic large language models (mini, Pro, Max) from Nex-AGI built for long-horizon tasks like computer use, web browsing, and autonomous coding, with Max built on a 1.6-trillion-parameter Mixture-of-Experts foundation.

  19. 22

    tpu-megakernels

    162Inferact/tpu-megakernels · Python

    A companion research repo with fused 'megakernel' decode implementations for two models on Google TPUs: Kimi K3 with DSpark speculative decoding, and Qwen3.8-27B with DFlash2 speculative decoding, written to accompany the blog post '700 TPS on Kimi K3: A Case for TPU Megakernels'.

Often tagged together