GitPop

AI & Machine Learning projects

LLMs, agents, inference runtimes, vector databases, and ML tooling — projects pushing the AI frontier.

36 projects tracked.

Browse by language

All →
Filter by language:Python22TypeScript5Go3C++2Rust2Jupyter Notebook1C1

Top 36 in AI & Machine Learning

PopScore over the last 7 day(s)

  • 1

    ggerganov/llama.cpp

    LLM inference in C/C++

    SlowingEnterprise
    Pop
    0
    C++#llm#inference#cpp
    71.8k10.4k
    AIMinimal-dependency C/C++ LLM inference engine with broad hardware support and aggressive quantization.expand
    Minimal-dependency C/C++ LLM inference engine with broad hardware support and aggressive quantization.
  • 4

    comfyanonymous/ComfyUI

    A powerful and modular stable diffusion GUI and backend.

    SlowingEnterprise
    Pop
    0
    Python#stable-diffusion#image-gen#ai
    56.3k5.8k
  • 5

    explosion/spacy

    Industrial-strength NLP in Python.

    SlowingEnterprise
    Pop
    0
    Python#nlp#ml#python
    30.4k4.5k
  • 6

    geekan/MetaGPT

    Multi-agent framework: GPT-driven software company simulation.

    SlowingEnterprise
    Pop
    0
    Python#agents#llm#multi-agent
    42.8k5.1k
  • 7

    ggerganov/whisper.cpp

    CPU-fast Whisper inference in plain C/C++.

    SlowingEnterprise
    Pop
    0
    C++#whisper#asr#cpp
    36.2k3.4k
  • 8

    gradio-app/gradio

    Build and share delightful machine learning apps, all in Python.

    SlowingEnterprise
    Pop
    0
    Python#dashboard#ml-ui#demos
    38.4k2.4k
  • 9

    huggingface/transformers

    State-of-the-art ML for PyTorch, TensorFlow, and JAX.

    SlowingEnterprise
    Pop
    0
    Python#llm#nlp#transformers
    132.0k26.5k
  • 10

    langchain-ai/langchain

    Build context-aware reasoning applications with LLMs.

    SlowingEnterprise
    Pop
    0
    Python#llm#agents#rag
    95.4k15.2k
  • 11

    langflowio/langflow

    Visual framework for building multi-agent and RAG applications.

    SlowingEnterprise
    Pop
    0
    Python#agents#llm#low-code
    28.5k4.3k
  • 12

    lllyasviel/Fooocus

    Focus on prompting and generating — minimal SD UI.

    SlowingEnterprise
    Pop
    0
    Python#image-gen#stable-diffusion
    42.5k3.2k
  • 13

    docling-project/docling

    Get your documents ready for gen AI

    HealthyEnterprise
    Pop
    0
    Python#ai#convert#documents
    65.0k4.8k
  • 14

    milvus-io/milvus

    A cloud-native vector database, storage for next-gen AI applications.

    SlowingEnterprise
    Pop
    0
    Go#vector-search#ai#similarity
    30.4k2.8k
  • 15

    mlflow/mlflow

    Open source platform for the machine learning lifecycle.

    SlowingEnterprise
    Pop
    0
    Python#mlops#tracking#experiments
    20.4k4.2k
  • 17

    openai/openai-python

    Official Python library for the OpenAI API.

    SlowingEnterprise
    Pop
    0
    Python#openai#llm#sdk
    22.4k3.6k
  • 18

    openai/whisper

    Robust speech recognition via large-scale weak supervision.

    SlowingEnterprise
    Pop
    0
    Python#asr#speech#openai
    78.5k8.4k
  • 19

    danny-avila/librechat

    Enhanced ChatGPT clone with multi-provider support, agents, and plugins.

    SlowingEnterprise
    Pop
    0
    TypeScript#chatgpt#ai#llm
    24.5k4.2k
    AISelf-hostable multi-provider ChatGPT-compatible UI with agents, plugins, and code execution.expand
    Self-hostable multi-provider ChatGPT-compatible UI with agents, plugins, and code execution.
  • 20

    qdrant/qdrant

    High-performance, massive-scale vector database.

    SlowingEnterprise
    Pop
    0
    Rust#vector-db#embeddings#search
    21.4k1.5k
  • 21

    ray-project/ray

    Ray is a unified framework for scaling AI and Python applications.

    SlowingEnterprise
    Pop
    0
    Python#distributed#ml#scaling
    35.4k5.8k
  • 22

    run-llama/llama_index

    Data framework for LLM applications — connect custom data to LLMs.

    SlowingEnterprise
    Pop
    0
    Python#rag#llm#framework
    38.5k5.2k
  • 23

    scikit-learn/scikit-learn

    Machine learning in Python.

    SlowingEnterprise
    Pop
    0
    Python#machine-learning#python#models
    60.4k25.2k
  • 24

    ollama/ollama

    Get up and running with large language models locally.

    SlowingEnterprise
    Pop
    0
    Go#llm#ai#local
    89.5k6.8k
    AIOne-binary local LLM runner with a simple `ollama run` UX and broad model support.expand
    One-binary local LLM runner with a simple `ollama run` UX and broad model support.
  • 25

    mendableai/firecrawl

    Turn entire websites into LLM-ready markdown or structured data.

    SlowingMature
    Pop
    0
    TypeScript#scraping#llm#crawler
    18.9k1.4k
    AIWeb scraper API that turns any site into clean markdown for LLM ingestion, with JS-rendering and anti-bot handling.expand
    Web scraper API that turns any site into clean markdown for LLM ingestion, with JS-rendering and anti-bot handling.
  • 26

    anthropics/claude-code

    Anthropic's official CLI for Claude.

    SlowingEnterprise
    Pop
    0
    TypeScript#claude#agent#cli
    18.6k1.1k
  • 27

    unslothai/unsloth

    2-5× faster LLM fine-tuning with less memory.

    SlowingMature
    Pop
    0
    Python#llm#fine-tuning#training
    18.2k1.4k
  • 28

    memvid/memvid

    Memory layer for AI Agents. Replace complex RAG pipelines with a serverless, single-file memory layer. Give your agents instant retrieval and long-term memory.

    SlowingMature
    Pop
    0
    Rust#ai#context#embedded
    16.4k1.2k
  • 29

    chroma-core/chroma

    Open-source embedding database for AI applications.

    SlowingMature
    Pop
    0
    Python#vector-db#embeddings#rag
    14.8k1.3k
  • 30

    pgvector/pgvector

    Open-source vector similarity search for Postgres.

    SlowingMature
    Pop
    0
    C#vector-search#postgres#extension
    14.2k580
  • 31

    weaviate/weaviate

    Weaviate is an open-source vector database.

    SlowingMature
    Pop
    0
    Go#vector-search#ai#semantic-search
    11.4k720
  • 32

    openai/openai-node

    Official Node.js / Deno library for the OpenAI API.

    SlowingEnterprise
    Pop
    0
    TypeScript#openai#llm#sdk
    8.9k1.2k
  • 33

    langchain-ai/langgraph

    Build stateful, multi-actor agent workflows with LangChain.

    SlowingMature
    Pop
    0
    Python#agents#llm#graph
    6.8k920
    AIGraph-based orchestration for stateful LLM agents with cycles, retries, and persistence.expand
    Graph-based orchestration for stateful LLM agents with cycles, retries, and persistence.
  • 34

    e2b-dev/e2b

    Open-source secure cloud sandboxes for AI agents.

    SlowingEnterprise
    Pop
    0
    TypeScript#agents#sandbox#llm
    6.4k580
    AIEphemeral firecracker sandboxes for AI agents — safe code execution in <200ms cold start.expand
    Ephemeral firecracker sandboxes for AI agents — safe code execution in <200ms cold start.
  • 35

    smol-ai/smol-course

    A hands-on course on aligning small language models.

    SlowingMature
    Pop
    0
    Jupyter Notebook#llm#course#alignment
    6.4k580
    AIHands-on, code-first course for aligning small LLMs (SFT + DPO + eval) on a single GPU.expand
    Hands-on, code-first course for aligning small LLMs (SFT + DPO + eval) on a single GPU.
  • 36

    lmsys/lmsys-chat-1m

    1M real LLM conversations with human preferences (research dataset).

    SlowingMature
    Pop
    0
    Python#dataset#llm#research
    3.8k380

How GitPop ranks projects

Read the full methodology
What is PopScore?
A 0–100 score combining star velocity (35%), fork engagement (20%), activity (25%), mentions (10%), and a noise filter (10%). Details →
Why not just use GitHub Trending?
Trending is dominated by awesome-* lists and tutorial repos. We weight commits and merged PRs above raw stars, and apply a noise penalty for content-only repos. Details →
How is the AI Snapshot generated?
An LLM summarizes each repo's README into strict JSON — TL;DR, 3 key features, difficulty rating. Cached and refreshed automatically. Details →