LLM: open-source projects on GitHub
2094 repositories on LLM, picked by the REDDYX AI agent. Each has a summary, key features and a breakdown. Sorted by stars.
// RELATED TOPICS
PENTEST HARNESS FOR HACKERS
Pentest Harness — a self‑hosted AI‑agent framework for authorized pentests, bug bounty, security labs, and CTFs. Connect your own model via API; all sessions run strictly
★ 471 · TypeScriptVectorRAG.Net — Vector RAG for .NET
High-performance library for semantic search and RAG on .NET. LSH + exact re-ranking via dot/cosine.
★ 470 · C#WRITING BENCHMARK – LOVED ELEMENTS
Tests how AI weaves 10 mandatory plot elements into a short story.
★ 470ASER: SELF-ASSEMBLY AI AGENT
Aser — lightweight framework for self-assembling AI agents on Python. Supports the a2a protocol, agente memory, and fast deployment via simple scripts.
★ 470 · PythonChorus: Multi-LLM code review
Chorus organizes collaborative code review with 2–4 LLMs before deployment. Use your CLI and Chorus will bring in other models for checking.
★ 469 · TypeScriptGUI AGENT RESOURCES
Curated collection of tools and frameworks for building GUI agents, simplifying work with interfaces.
★ 469THREAT MODELING FOR AI AGENTS
grim-prompter points Claude Code, Codex, or any agent at your repo and delivers an evidence-traced threat model. Covers 36 architectures and 29 methods (LINDDUN, STRIDE).
★ 469AI ASSISTANT FOR SUBLIME TEXT
Integration of GPT‑5, Opus 4.6, Gemini 3 and Ollama in Sublime Text for instant code and autocompletion.
★ 467 · PythonAVIBE - Your Local AI Agent
Avibe is a local OS for AI agents that lets you run Claude Code, Codex, and OpenCode directly in your browser or any chat app.
★ 466 · PythonAGENT-STYLE: 21 RULES
agent-style provides 21 style rules for AI agent code and text. Connect it to Claude, Codex, Copilot, Cursor, or Aider — and the output becomes professional.
★ 465 · PythonCore AI Model Collection
A model zoo for Apple Core AI with Qwen3.5 and Gemma 4 conversion, custom Metal kernels, and a Swift runner.
★ 465 · SwiftEXXPERTS LOCAL AI AGENT
An AI agent that runs entirely on your device, with controllable memory and built‑in web search.
★ 465 · TypeScriptTELEMEM - MEM0 KILLER IS HERE
High-performance replacement for Mem0 with semantic deduplication, long-term dialog memory, and multimodal video analysis.
★ 464 · PythonAGENTGG: AI-POWERED SAST SCANNER
Open-source agentic SAST engine behind hundreds of published zero‑days. Over 100 AI agents analyze any repository or PR diff, supporting the use of your own model.
★ 464 · TypeScriptSwarmVault: Local Wiki for LLM
Open tool for building knowledge graphs, RAG database, and agent memory. An Obsidian alternative for personal knowledge management and AI second brain.
★ 463 · TypeScriptAGENTS: AI Agent for Data
Agents — a toolkit for creating AI agents that automate data engineering. It supports ETL pipelines, LLM, and workflow management.
★ 463 · PythonFASTEST AI GATEWAY ON RUST
Antirion — open-source AI gateway in a single Rust binary. Routes, secures, budgets and observes every LLM call: OpenAI, Anthropic, Gemini, Bedrock. Built-in management c
★ 462 · RustSPEAKEASY – SDK Generator
Speakeasy generates type-safe SDKs, Terraform providers, and CLI from OpenAPI, simplifying API integration.
★ 461 · JavaScriptAGENT - Agent AI for macOS
An AI agent hub for Mac: computer use, automation, and coding. Support for 18+ LLMs.
★ 460 · SwiftGet Ready for ML Interview
A collection of ML questions and answers for interview preparation. A quick checklist to help you stand out.
★ 458 · MarkdownSIMULATOR-CONTROLLER - VIRTUAL AI PIT-CREW
Using GPT, Simulator-Controller creates a realistic virtual pit-crew for racing simulators, giving the feeling of a real team on the track.
★ 458 · AutoHotkeyEVOLVING AGENT FOR TEAMS
The director agent studies the team's work via CLAUDE.md and linked memory statements, forms a plan and delegates tasks to Claude Code or local models. Lessons are stored
★ 458 · TypeScriptGOBII-PLATFORM: AI Worker 24/7
gobii-platform — your always-on AI employee in Python, with 456 stars on GitHub, easy to set up and ready to work.
★ 456 · PythonOpenRCA: AI Finds Root Causes in Code
OpenRCA is an LLM-based framework for automatic root cause detection of failures in software. Accepted at ICLR 2025.
★ 453 · PythonHYPHA Agent Framework
TypeScript framework for production‑grade LLM agents. Enables rapid creation and management of agents via a flexible harness architecture.
★ 452 · TypeScriptAI AGENTS ON AZURE AND GPU
Repository with code, benchmarks, and guides for AI agents and Deep Learning. Deployment on Microsoft Foundry and Azure, fine-tuning, inference optimization, and GPU test
★ 451 · Jupyter NotebookSMITHERS - KEEP CODE UNDER CONTROL
Smithers ensures reliable execution of agent code: it survives failures, waits for human confirmation, and rolls back errors. It supports Claude Code, Codex, Gemini, and
★ 449 · JavaScriptSANDBOXED.SH - SAFE ENVIRONMENT FOR AI
Isolated sandboxes, a skills library, encrypted secrets, and OKX security checks for autonomous on-chain AI agents.
★ 446 · RustRUNPROMPT: LLM PROMPTS FROM TERMINAL
Run LLM prompts directly from the command line. Simple, fast and convenient.
★ 446 · PythonSIMVYN: MANAGE ALL DEVICES
simvyn — a universal devtool for iOS, Android, and real devices: control them via a unified panel and CLI.
★ 446 · TypeScriptHARNESS-RSI: AGENT SELF-EVOLUTION & EVALUATION
Harness-RSI is a bibliographic project on recursive AI agent self-improvement. Contains theory, evaluation methods, evidence, and bilingual reading list. 431 GitHub stars
★ 446 · TeXLYNKR SIMPLIFYING CODE WORK
Lynkr is a CLI utility that acts as an HTTP proxy for fast interaction with Claude Code and streamlines development workflows.
★ 445 · JavaScriptONECOMPRESSION: Lossless LLM Compression
The Python package OneCompression simplifies compression of large language models while preserving accuracy and speeding up inference.
★ 445 · PythonPOCKETRISU - SELF‑BUILT AI CHAT
PocketRisu is a self‑hosted platform for AI role‑play chats, a fork of Risuai that runs on your PC or personal server.
★ 444 · TypeScriptPLATFORM FOR AI AGENTS
The open CAIPE platform helps create, deploy, and manage AI agents and their workflows on TypeScript.
★ 444 · TypeScriptOpenHack Security Agent
OpenHack is an open-source Python security agent that automatically finds vulnerabilities and suggests patches.
★ 443 · PythonARTEMIS AI COMPANION DESKTOP
ARTEMIS is an interactive AI companion with a live 2D model that generates images, answers questions and holds a conversation via GPT‑Sovits. Integrated with ComfyUI for
★ 443 · PythonULTRACONTEXT FOR AI AGENTS
Ultracontext — open source context infrastructure for AI agents. Automatically captures and shares agent context everywhere.
★ 442 · JavaScriptRLLM: Learning Tables with LLM
A PyTorch library for training relational tables using large language models.
★ 441 · PythonPORTABLE MEMORY FOR AGENTS
Portable semantic memory for AI agents: core, TS‑SDK, framework adapters, MCP server, CLI, and plugins.
★ 440 · TypeScriptMIXAMO-LLM-MOCAP: VIDEO TO ANIMATION
The project turns any video into a Mixamo‑rig animation using GVHMR pose estimation, spec‑driven retargeting, and FK application in Blender via MCP. The entire pipeline i
★ 440 · PythonQODEASSIST: AI Qt Creator Assistant
QodeAssist is an AI plugin for Qt Creator that suggests code, generates templates, and speeds up C++ development.
★ 439 · C++TEXT CLASSIFICATION WITHOUT KEY
Zero-shot text classification over plain HTTP — no API key, no account. One Cloudflare Worker, CLI and MCP server work together. Try it now: https://classifier.dev
★ 439 · TypeScriptMARM-MEMORY: LOCAL AI MEMORY
Memory layer for Claude Code and Cursor: combines session history, code indexes, and concept graphs in SQLite, providing cloud‑free context.
★ 438 · PythonNEUROARXIV: AI CHECKS ARXIV BEFORE CODING
A skill that eliminates the need to write code from scratch: Claude searches real papers on arXiv before proposing a new architecture. This reduces the risk of reinventin
★ 438 · TypeScriptABC-GRPO: ALL-QUADRANT BOUNDING MODEL
ABC-GRPO implements all-quadrant bounded clipping for GRPO, improving LLM training stability. Code based on arXiv:2601.03895, written in Python. Project has 442 GitHub st
★ 438 · PythonPREDICT-RLM Self-Governing LM
An LM runtime where the model invokes submodels via DSPy signatures. You get transparent trajectories and growing performance.
★ 437 · PythonREDIS-VL-PYTHON - VECTOR LIBRARY FOR AI
RedisVL is a Python client for working with vector data in Redis, simplifying the creation of AI applications and similarity search.
★ 437 · PythonLLM AGENT HARNESS OVERVIEW
Overview of over 110 papers and 23 systems for building agents for large language models.
★ 437Xiryn Personal AI Agent
Open‑source personal agent that runs isolated Linux desktops in Docker, supports any LLM with your keys, and offers an encrypted vault controllable via Telegram.
★ 435 · TypeScriptEDGECHAINS: GENAI IN JAVASCRIPT
Full‑stack library for working with AI. Prompts and chains are managed declaratively via jsonnet, not hidden in classes.
★ 434 · JavaScriptDEVELOPER DOCUMENTATION
AnythingLLM — AI platform for working with documents, YouTube, and hosting resources. Boost your ideas!
★ 433 · MDXCharDesk: Draw Diagrams with Code
CharDesk lets you draw diagrams, interfaces, and slides as editable text using a Unicode canvas. It works for both humans and AI agents.
★ 433 · TypeScriptPRYSAI PLAYBOOK: COMPLETE LLM GUIDE
LLM playbook on practice: 8 locales, unified core, Codex track and adapters for ChatGPT, Claude Code, Gemini, DeepSeek, Grok.
★ 433 · PythonCLI Agent Orchestration
Audit-grade orchestration of multi-agent CLI tools with HMAC logs, agent signatures, and artifact lineage — compliance-ready.
★ 432 · PythonPI-BOOK - CREATE AGENT WITH VUE
The pi-book project offers architectural notes with source code for creating an AI agent on Vue.
★ 432 · VuePLUGIN LIST FOR DEEPSEEK HARNESS
A curated list of plugins, skills, MCP servers, patch/profile layers for DeepSeek Harness. Includes visualization tools,PPT generation,coding,agents, and auto‑research lo
★ 432 · TypeScriptEXPERIMENTAL INFERENCE LIBRARY
resin — experimental Python library for building LLM inference engines. Supports CUDA, GPU acceleration, and models like DeepSeek and GPT-OSS. Early-stage project with 25
★ 432 · PythonMCPTOON: SAVES 99% OF MCP TOKENS
Standard MCP discovery burns 71,929 tokens. mcptoon drops it to 581 — 99.2% savings. Zero dependencies.
★ 431 · HTMLMODELS - AI MODEL BROWSER
TUI and CLI for browsing AI models, benchmarks, and providers directly from the terminal in Rust.
★ 430 · RustNew projects on this topic in Telegram
Telegram →