LLM: open-source projects on GitHub
2095 repositories on LLM, picked by the REDDYX AI agent. Each has a summary, key features and a breakdown. Sorted by stars.
// RELATED TOPICS
LM RING: COMPARE AND VOTE
Open self-contained arena for LLM comparison: vote for the best models and watch the leaderboard in real time.
★ 208 · TypeScriptAGENTUSE: AI AGENT AUTOPiLOT
Agentuse lets you run AI agents of any scale locally, on a cron schedule, in CI/CD pipelines, or inside Docker containers. It supports any LLM model and is written in Typ
★ 208 · TypeScriptFULL GUIDE TO LLM EVALUATION
LLMEvaluation provides a full guide to LLM evaluation methods aiding selection of optimal techniques for specific use-cases. It contains a metrics taxonomy, step‑by‑ste
★ 208 · HTMLHERMES MEMORY INSTALLER TOOL
Production‑grade memory sidecar for AI agents: integrates gbrain, Hindsight and 3‑tier recall. Agent‑agnostic, battle‑tested, works with Hermes, Claude, Cursor and any LL
★ 208 · PythonGIT FOR INTENT OF AGENT CODE
Plugin for AI coding agents that records every dialogue as a 'thought commit'. Links code commits to reasons for changes, allowing `cogit why` to find justification for a
★ 208 · TypeScriptJDOCMUNCH-MCP: EFFICIENT MCP SERVER
Token-efficient server for searching and extracting documentation through structured indexing of sections. Written in Python, already 200+ stars.
★ 207 · PythonAI IN BROWSER WITHOUT SERVER
Library of providers for Vercel AI SDK runs models directly in the browser. Simplifies client-side AI integration and unifies work across vendors. 205 stars on GitHub.
★ 207 · TypeScriptMONGODB LENS: MCP SERVER
MongoDB Lens is a full-featured MCP server for MongoDB databases. Allows LLMs to read schemas, execute queries, and manage data via Model Context Protocol.
★ 207 · JavaScriptREKAL — MEMORY FOR AI AGENTS
Your AI agent starts every session blank. Rekal adds long-term memory: preserves code context, decisions, and patterns across sessions. Go CLI integrates with Claude Code
★ 206 · GoRELIABLE AI AGENT ON PYTHON
Protect personal data with local request scrubbing via a local LLM before sending to the cloud.
★ 205 · PythonCONNECT AI TO UNITY AND GODOT
glade-mcp connects any MCP‑compatible AI client (Claude Code, Cursor, Windsurf) to Unity or Godot. It provides 235+ tools, an editor‑aware prompt, GDD context, semantic c
★ 205 · C#UNIVERSAL AI AGENT IN SINGLE BINARY
Coddy-agent — autonomous agent in static Go binary. TUI, ACP server, OpenAI-compatible API with web UI, Telegram gateway, cron, swarm-relay. MCP, long-term memory, contex
★ 205 · GoDaoCloud Docs: Official Documentation
Official documentation for DaoCloud Enterprise 5.0 in Python: setup, deployment, and management of the enterprise cloud platform.
★ 204 · PythonCAI - HASSLE-FREE AI RESULTS
cai is a simple CLI tool in Rust that automates LLM work: set a goal, get a result without prompts.
★ 204 · RustCLAUDECTL AGENT ORCHESTRATION
claudectl orchestrates a swarm of Claude Code agents with a local ‘brain’ that follows your preferences. Written in Rust, it already has 202★ on GitHub.
★ 204 · RustEUROEVAL: EUROPEAN LLM BENCHMARK
EuroEval — framework for evaluating LLMs on European languages. Supports 20+ languages, automatic metrics, and public leaderboard. Written in Python.
★ 204 · PythonLITER-LLM UNIVERSAL LLM CLIENT
The client supports 142+ providers and 11 native libraries, built on a fast Rust core.
★ 203 · RustMYTHOS-ROUTER: LOCAL AI-AGENT
Local router with a writing discipline for precise coding via LLM.
★ 203 · TypeScriptAI-DIAL SDK: Build AI Apps
Framework for quickly building apps and model adapters for AI DIAL in Python. Supports integration with LLM and RAG.
★ 203 · PythonFLOWCRAFT - AI AGENTS IN GO
Production-grade Go SDK for building AI agents with long-term memory, knowledge search, and voice control.
★ 202 · GoGENERATE AI WORKFLOW FROM TEXT
FlowX is an MCP server and Python SDK to generate, run, and refine agent workflows from natural language. Supports A2A and FastAPI.
★ 202 · PythonDEEPSEEK-CODE-WHALE - AI CODING AGENT
Terminal AI agent for DeepSeek with MCP integration, skill system, and cache optimization for a 97% hit rate.
★ 201 · GoONIONCLAW: TOR FOR AI AGENTS
Provides AI agents full access to the Tor network and dark web data without configuration.
★ 201 · PythonNODE9-PROXY: SHIELD FOR AI AGENTS
node9-proxy provides deterministic management and auditing of autonomous AI agents, ensuring safe code execution.
★ 201 · TypeScriptBitemporal AI Memory System
Bitemporal memory system with six-variable decay, time-graded recall, TTL expiry, deduplication, and competitive retrieval using Neo4j, PostgreSQL, and Redis.
★ 201 · PythonGLYD: LOSSLESS AI COMPRESSION
Glyd reduces GPU memory by 33% without losing data, ensuring bit‑for‑bit identity. The project is written in Rust and supports CUDA, C‑ABI, and CLI for convenient benchma
★ 200 · RustSuperLog: observability and debugging
SuperLog is an open-source TypeScript observability tool that automatically detects and fixes errors in code. It integrates with CI/CD and provides real-time application
★ 198 · TypeScriptSelf-Hosted AI Agent Platform
Open-source, self-hosted agent platform using ReAct and Harness patterns, with AI customer support as the reference application.
★ 198 · PythonNOOBOT CHEAP AI AGENT
Self‑hosted workspace for AI agents with low cost: support for tool calling, MCP, multi‑model routing and sandboxed execution. Enables creating multi‑agent workflows and
★ 195 · JavaScriptNexus Agent Context
Nexus is a shared context plane where AI agents and humans collaborate, exchange knowledge, and grow together.
★ 194 · PythonBeginner AI Learning Path
A step-by-step AI roadmap covering foundations to production with free resources and hands-on mini-projects for each layer.
★ 194DeepSeek Local Setup Guide
A Russian‑language guide that walks you through installing and running DeepSeek models (V4, R1) locally on PC, Mac, or server using tools like Ollama, llama.cpp, LM Studi
★ 192 · ShellPersona SDK for LLM Apps
Defines persona cards, loads configurations, and adapts prompts across LLM providers like OpenAI and Anthropic.
★ 187 · TypeScriptMULTI-CLI - MULTI-PROFILE MODE
Run multiple AI coding tool accounts simultaneously, saving time on switching.
★ 181 · PowerShellMUXI Infrastructure for AI
An open platform for deploying AI agents in production. Simple integration, scalability, and monitoring.
★ 179Comprehensive AI Engineering Course
Free 173-lesson course covering data, ML, deep learning, and AI engineering with 166 runnable labs. Covers math, code, systems, and security under CC BY-NC 4.0.
★ 175 · PythonPRODUCT‑MODE FOR AI TEAMS
The project provides a CLAUDE.md/AGENTS.md template with principles of problem framing, scope, tradeoffs, outcome measurement for PM, eng team. It works as a drop‑in file
★ 171 · TypeScriptGODCODER — LOCAL AI AGENT
Open cross-platform agent for code. Your code stays on your machine, LLM keys are yours.
★ 165 · RustKVarN: KV-Cache Quantization
KVarN is a native KV-cache quantization backend for vLLM, increasing context by 3-5x, raising throughput above FP16 and maintaining FP16 accuracy without calibration, act
★ 159 · PythonModel Benchmarking Platform
The project provides tools for comparing machine models as real-time controllers of engineering loops, measuring their latency, stability, and resource consumption.
★ 155 · PythonAdaptive Agent Inference Framework
EvolvingInfra dynamically optimizes inference for diverse agent workloads, adapting scheduling and resource allocation in real-time.
★ 154 · PythonAGENTSMITH: Universal AI Harness
agentsmith — lightweight core + task profiles for any AI agent, assembled with a single shell script.
★ 152 · ShellLoRA Fine-Tuning Sprint
Project accelerates LoRA fine-tuning by fixing task and hardware, and posts results to a public wall-clock leaderboard based on modified nanogpt.
★ 151 · PythonSMART PLAYER IN MINECRAFT
AI agent plays Minecraft by itself! Server‑side Fabric mod 1.21.3 — LLM plans actions, bot mines, builds and survives. ⭐211
★ 149 · JavaLOOMFLOW: LIGHTWEIGHT AI WORKFLOW BUILDER
LoomFlow is a lightweight AI-native workflow builder. Describe your idea in natural language, get a visual scheme and immediately publish it as an API. Install with one c
★ 138 · TypeScriptAI Short Video Pipeline
Fully automated pipeline for creating short videos with AI: FastAPI orchestration, Spring Boot gateway with fault tolerance, metrics and observability, and quality contro
★ 137 · PythonGoal-Flow Graph Agent
Goal-Flow is a LangGraph framework that combines workflow graphs and agent loops, translating Dify DSL into executable code and switching Dify/OpenAI protocols.
★ 137 · PythonSchema-Aware SQL Access Control
Filters database schemas by user permissions before LLM queries, enforcing row-level security across Postgres, Oracle, MySQL, and SQL Server. Cuts prompt tokens by 75%.
★ 137 · PythonMCP Server for Tennis
The MCP server provides live scores, odds, and win probability to LLM agents for tennis.
★ 135 · JavaScriptOn-Demand Model Alignment
Aligns LLMs instantly per request for verified users or always-on mode. Integrates with Hermes, OpenClaw, and custom harnesses.
★ 126 · PythonArchitectural Patterns for AI Systems
Catalog of 21 architecture patterns (RAG, agents, vector DBs, inference) with a tutorial and links to open-source prototypes. In Russian and Chinese.
★ 125 · VueCurated Self-Hosted AI Tools
Weekly health-checked list of open-source AI for local LLMs, chat UIs, agents, RAG, and media generation. Dead projects auto-removed.
★ 124 · JavaScriptHands-On LLM Full-Stack Course
Comprehensive tutorial implementing LLMs from scratch: builds Transformer in PyTorch, reproduces Pretrain/SFT/LoRA/DPO/GRPO on single GPU, covers PEFT/Agent/RAG with real
★ 124 · Pythonnl-veil: AI CODER IN ZIG
nl-veil — open-source AI coding assistant with parallel agents and persistent project memory. Works with local models, Ollama or cloud LLMs on Windows, macOS and Linux. W
★ 113 · ZigClaude Code Backdoor
The project lets you run Claude Code on any AI provider: DeepSeek, Groq, Ollama, and OpenRouter through a single Python interface.
★ 109 · PythonLLM TCO Calculator
Bindwidth estimates hardware requirements and TCO for running LLMs locally, using real data to calculate required GPU/CPU and energy consumption, depreciation, and licens
★ 109 · JavaScriptBhagavad Gita Knowledge Graph
Explore all 700 verses in 3D with 20 commentaries, audio recitations, GraphRAG chat, and private journal — fully offline.
★ 106 · CypherORCHESTRATING CLAUDE AGENTS
Mavericks-os orchestrates Claude Code sub-agents in long-running tasks. Stores state in artifacts, not chat history. Supports multi-repo, dependencies, and QA gates.
★ 105 · JavaScriptDeepFake AI Kit
LLM-based toolkit for detection, analysis, and responsible creation of synthetic media in real time.
★ 99 · C#Task-Oriented AI Workbench
An AI workbench where agents autonomously select tools like search, web reading, maps, and knowledge bases in a single loop, with fully traceable sources.
★ 99 · PythonNew projects on this topic in Telegram
Telegram →