anrgct

duckdb-ai. Enhance DuckDB with AI functions, supporting all providers as well as local models

programasweights-python. Python SDK for ProgramAsWeights — compile natural language specs into neural programs that run locally

MemGUI-Agent. Official code for "MemGUI-Agent: An End-to-End Long-Horizon Mobile GUI Agent with Proactive Context Management"

mastra. Mastra is the modern TypeScript framework for AI-powered applications and agents.

pon. Python 3.14, compiled to metal — JIT & AoT native compiler and runtime in Rust. Cranelift backend, ruff parser, Green Tea GC, byte-exact differential testing against CPython.

ternlight. Python

dinotty. Multi-device terminal server for AI coding agents. Server-side VTE, session persistence, file browser, web preview, plugin system. Self-hosted.

lisette. A little language inspired by Rust that compiles to Go

turso. Turso is an in-process SQL database, compatible with SQLite.

eve. The Framework for Building Agents

lotus. Optimized Agentic and LLM Bulk Processing Over Your Data

haystack. Open-source AI orchestration framework for building context-engineered, production-ready LLM applications. Design modular pipelines and agent workflows with explicit control over retrieval, routing, memory, and generation. Built for scalable agents, RAG, multimodal applications, semantic search, and conversational systems.

liteparse. A fast, helpful, and open-source document parser

terax-ai. Lightweight (7MB) Terminal-first AI-native dev workspace

chapi. CHAPI (Common Hierarchical Abstract Parser and Information Converter) streamlines code analysis by converting diverse language source code into a unified abstract model, simplifying cross-language development. Chapi 是一个通用层次抽象解析器与信息转换器，它可以将不同编程语言的源代码转换为统一的层次抽象模型。

model2vec. Fast State-of-the-Art Static Embeddings

agentsview. Local-first session search, analytics, insights, and token use statistics for coding agents, supporting Claude Code, Codex, and more than 20 other agents.

OpenMythos. A theoretical reconstruction of the Claude Mythos architecture, built from first principles using the available research literature.

flue. The sandbox agent framework.

codebase-memory-mcp. High-performance code intelligence MCP server. Indexes codebases into a persistent knowledge graph — average repo in milliseconds. 158 languages, sub-ms queries, 99% fewer tokens. Single static binary, zero dependencies.

Trellis. The best agent harness.

ASCILINE. A high-performance, real-time ASCII video rendering engine. Streams binary-encoded frames via WebSockets for ultra-low latency, 30 FPS playback using HTML5 Canvas and requestAnimationFrame.

ratty. A GPU-rendered terminal emulator with inline 3D graphics 🐀🧀

OminiX-MLX. MLX implementation of OminiX for LLM, image generataion, ASR and TTS

3D-Speaker. A Repository for Single- and Multi-modal Speaker Verification, Speaker Recognition and Speaker Diarization

oxc. ⚓ A collection of high-performance JavaScript tools.

ast-grep. ⚡A CLI tool for code structural search, lint and rewriting. Written in Rust

codeql. CodeQL: the libraries and queries that power security researchers around the world, as well as code scanning in GitHub Advanced Security

club-3090. Community recipes for serving LLMs on RTX 3090/4090/5090 CUDA gpus. Multi-engine (vLLM, llama.cpp, ik_llama) and model-agnostic. Currently shipping Qwen3.6-27B Qwen3.6 35B Gemma 4 26B Gemma 4 31B configs for 1× and 2× cards.

perry. A native TypeScript compiler written in Rust. Compiles TypeScript directly to executables using SWC and LLVM.

robrain. Shared memory across your team and your AI agents — with judgment about what's worth keeping.

oh-my-pi. ⌥ AI Coding agent for the terminal — hash-anchored edits, optimized tool harness, LSP, Python, browser, subagents, and more

container. A tool for creating and running Linux containers using lightweight virtual machines on a Mac. It is written in Swift, and optimized for Apple silicon.

openwolf. Sharper context. Fewer tokens. Open-source middleware for Claude Code.

forkd. Fork() for AI agent microVMs. Spawn 100 children in ~100ms from a warm parent; BRANCH a live VM in ~150ms. KVM-isolated, snapshot CoW.

MiMo-Code. MiMo Code: Where Models and Agents Co-Evolve

tokenplay. tokenplay- 玩转Token，Token照妖镜，纯度检测器！

HippoRAG. [NeurIPS'24] HippoRAG is a novel RAG framework inspired by human long-term memory that enables LLMs to continuously integrate knowledge across external documents. RAG + Knowledge Graphs + Personalized PageRank.

amla-sandbox. Python

LCLM. latent context language models

vLLM-2080Ti-Definitive. The definitive vLLM runtime for dual RTX 2080 Ti 22GB + NVLink, delivering Qwen 27B local inference with 100+ tok/s single-request decode with support of FP8 weight

CoSIL. [ASE 2025] CoSIL: Issue Localization via Iteritive Code Graph Searching

duckle. Open-source local-first ETL/ELT studio: visual data pipelines, 290+ connectors, dbt and on-device AI, run on DuckDB.

turbovec. A vector index built on TurboQuant, written in Rust with Python bindings

headroom. Compress tool outputs, logs, files, and RAG chunks before they reach the LLM. 60-95% fewer tokens, same answers. Library, proxy, MCP server.

openclaw-dashboard. 🔐 Secure, real-time monitoring dashboard for OpenClaw AI agents. Auth, TOTP MFA, cost tracking, live feed, memory browser and more.

ruflo. 🌊 The leading agent meta-harness. Deploy intelligent multi-player swarms, coordinate autonomous workflows, and build conversational AI systems. Features adaptive memory, self-learning intelligence, RAG integration, and native Claude Code / Codex / Hermes and many more Integrated

edulab. HTML

dbx. 20MB, lightweight, cross-platform database client. Supports MySQL, PostgreSQL, SQLite, Redis, MongoDB, DuckDB, ClickHouse, SQL Server and more.20MB，轻量级跨平台数据库客户端、数据库管理工具。支持 MySQL、PostgreSQL、SQLite、Redis、MongoDB、DuckDB、ClickHouse、SQL Server 等。

CubeSandbox. Instant, Concurrent, Secure & Lightweight Sandbox for AI Agents.

llm2vec. Code for 'LLM2Vec: Large Language Models Are Secretly Powerful Text Encoders'

llm2vec-gen. Code for `LLM2VEC-GEN: Generative Embeddings from Large Language Models`

rmux. Universal Rust multiplexer with a typed SDK — drive any CLI or TUI app from code. Native on Linux, macOS, and Windows.

semble. Fast and Accurate Code Search for Agents. Uses ~98% fewer tokens than grep+read

cocoindex-code. A super light-weight embedded code search engine CLI (AST based) that just works - improves speed and efficiency for coding agent 🌟 Star if you like it!

ds4. DeepSeek 4 Flash and PRO local inference engine for Metal, CUDA and ROCm

rf-detr.cpp. rt-detr implementation in C++ with ggml

embedding-benchmark. A simple benchmark for comparing embedding models, dimensions, and retrieval quality across public datasets using RedisVL, OpenAI, Ollama, and NanoBEIR.

optillm. Optimizing inference proxy for LLMs

EnterpriseRAG-Bench. Dataset and benchmark for RAG on company internal documents.

blockify-agentic-data-optimization. Blockify: Agentic data optimization for enterprise RAG and agentic search. Blockify replaces naive chunking with deduplicated IdeaBlocks — ≈78X accuracy aggregate improvement, 2.29X vector search accuracy, 40X compression, 3.09X token efficiency. Integrates with LlamaIndex, LangChain, Obsidian, Milvus, Elastic, Supabase, Cloudflare.

gguf-surgeon. Surgical metadata edits for GGUF model files — CLI, TUI, JSON patches, schema validation, atomic crash-safe writes.

gguf-editor. gguf-editor - a vscode extension

instrument-agnostic-amt. Neural Semi-CRF based Automatic Music Transcription

DeepSeek-Reasonix. DeepSeek-native AI coding agent for your terminal. Engineered around prefix-cache stability — leave it running.

node-llama-cpp. Run AI models locally on your machine with node.js bindings for llama.cpp. Enforce a JSON schema on the model output on the generation level

MiniCPM-Desk-Pet. a local-first desktop pet powered by MiniCPM5

Bonsai-Image-Demo. Generate images locally

X2SAM. [ECCV2026] X2SAM: Any Segmentation in Images and Videos

AstrAI. A lightweight Transformer training & inference framework

lark-coding-agent-bridge. Bot that bridges Feishu/Lark messenger with a local Claude Code or Codex CLI. Streaming cards, per-chat sessions, multiple workspaces

Proma. 把最丝滑的通用 Agent 体验带进你的工作流，为 100x 专业用户而生的未来产品，正在实现 proactive Agent 阶段。基于 Claude Agent SDK 的完整开源实践，原生支持飞书群聊调用、灵活接入任意大模型供应商 —— 让顶级 Agent 能力真正跑在你每天用的地方。

opensquilla. OpenSquilla — Token-Efficient AI Agent with same budget, higher intelligence density

LatentMAS. [ICML 2026 Spotlight] Latent Collaboration in Multi-Agent Systems

forge. A Python framework for self-hosted LLM tool-calling and multi-step agentic workflows

Q-RAG. Official repository for the ICLR 2026 Oral Paper🔥 “Q-RAG: Long Context Multi-Step Retrieval via Value-Based Embedder Training”

HRM-Text. HRM-Text is a 1B text generation model based on the HRM architecture, strengthened by task completion and latent space reasoning.

mlx. MLX: An array framework for Apple silicon

sapiens2. 1K resolution vision transformers pretrained on 1B human images.

OpenUtau. Open singing synthesis platform / Open source UTAU successor

signal. Online MIDI Editor: signal

open_provence. ✂️ OpenProvence: Open-Source, Efficient, and Robust Context Pruning for Retrieval-Augmented Generation

embedding_rerank_retrieval. 本项目是针对RAG中的Retrieve阶段的召回技术及算法效果所做评估实验。使用主体框架为LlamaIndex.

jina-grep-cli. Semantic grep powered by Jina embeddings v5 (MLX on Apple Silicon)

qdrant. Qdrant - High-performance, massive-scale Vector Database and Vector Search Engine for the next generation of AI. Also available in the cloud https://cloud.qdrant.io/

guppylm. A ~9M parameter LLM that talks like a small fish.

Grinta-Coding-Agent. Local-first autonomous coding agent that plans, executes, validates, and finishes software tasks end-to-end.

nanowhale. Python

mirage. A Unified Virtual Filesystem For AI Agents

Rocket.Chat. The Secure CommsOS™ for mission-critical operations

multica. The open-source managed agents platform. Turn coding agents into real teammates — assign tasks, track progress, compound skills.

sam-3d-objects. SAM 3D Objects

sam-3d-body. The repository provides code for running inference with the SAM 3D Body Model (3DB), links for downloading the trained model checkpoints and datasets, and example notebooks that show how to use the model.

unsloth. Unsloth Studio is a web UI for training and running open models like Gemma 4, Qwen3.6, DeepSeek, gpt-oss locally.

minimind-o. 🎙️ 「大模型」从0训练0.1B能听能说能看的全模态Omni模型！A 0.1B Omni model trained from scratch, capable of listening, speaking, and seeing!

X-AnyLabeling. Effortless data labeling with AI support from Segment Anything and other awesome models.

ComfyUI-Easy-Use. In order to make it easier to use the ComfyUI, I have made some optimizations and integrations to some commonly used nodes.

devpiano. Windows 虚拟 MIDI 钢琴，支持 VSTi、演奏和视频的录制与导出。| Windows virtual MIDI piano, supports VSTi, performance, and recording and exporting of videos.

structured-cot. Structured Chain-of-Thought

qwen36-dual-3090. Qwen3.6-27B on dual RTX 3090 — TP=2 recipe, vLLM nightly, MTP + fp8 KV, validated for concurrent serving

Fun-ASR-GGUF. 用 onnx 和 gguf 格式混合运行 Fun-ASR-Nano 模型全流程

lucebox. Fast LLM speculative inference server for consumer hardware.

voicebox. The open-source AI voice studio. Clone, dictate, create.

CanvasMind. A modern low-code visual programming IDE built on NodeGraphQt and qfluentwidgets, supporting drag-and-drop component orchestration, asynchronous execution, file operations, loop control, and one-click export of workflows into standalone runnable projects—enabling seamless transition from development to deployment.

AI-CanvasPro. AI-CanvasPro

Subtitle-Translation. Subtitle-TranslationA 是一款专业的视频字幕翻译工具，利用最新的Whisper-CD实现高精度的语音识别。

mlx-qwen3-asr. Qwen3-ASR speech recognition on Apple Silicon via MLX

transby2. Python

claude-agent-acp. Use Claude Agent SDK from any ACP client

Foundry-Local. C++

dflash-mlx. Lossless DFlash speculative decoding for MLX on Apple Silicon

ReadAny. TypeScript

Comfyui-zhenzhen. https://ai.t8star.org/register?aff=dP7j Comfyui的各种平价API调用节点，支持seedance2.0（满血版，支持真人，不排队）

freemocap. Free Motion Capture for Everyone 💀✨

flash-moe. Running a big model on a small laptop

Kinema. TypeScript

Falcon-Perception. Inference repo for Falcon-Perception and Falcon-OCR model, early-fusion, natively multimodal, dense Autoregressive Transformer models.

Tapnow-Studio-PP. 设计师随心所欲的一枚无限画布

GVHMR. Code for "GVHMR: World-Grounded Human Motion Recovery via Gravity-View Coordinates", Siggraph Asia 2024, TPAMI 2026

listen1_chrome_extension. one for all free music in china (chrome extension, also works for firefox)

ComfyUI. The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface.

ChatDev. ChatDev 2.0: Dev All through LLM-powered Multi-Agent Collaboration

Tapnow-Studio--. 守护设计师的肝和钱包

reze-studio. Web-based MMD animation curve editor

claude-code-source. Claude Code complete source restoration from npm package — all native modules, no features cut

ColorTxt. 「彩读 2.0 AI+」——一款会给内容上色的本地 TXT 小说阅读器，带给你不一样的阅读体验！（也支持打开常见的电子书格式，如 .epub）。还有章节识别、简繁互转、划线标注、记笔记、多角色语音朗读、AI 阅读助手、AI 智能排版等功能。支持 macOS、Windows 和 Linux。

threejs-editor-beta. Three.js lowcode editor beta

ACE-Step-1.5. The most powerful local music generation model that outperforms almost all commercial alternatives, supporting Mac, AMD, Intel, and CUDA devices.

claude-code-rust. 🚀 Rust 全量重构的 Claude Code - 性能提升 2.5x，体积减少 97% | High-performance Rust implementation of Claude Code with 2.5x faster startup and 97% smaller binary

Claude-Code. 可运行的Claude Code源码

agentflow. Orchestrate thousands of agents and harnesses as a graph programatically

claw-code. An agent-managed museum exhibit, built in Rust with Gajae-Code / LazyCodex — developed and maintained with no human intervention.

Infernux. Open-source game engine with a C++17 / Vulkan runtime and a Python production layer.

pretext. Fast, accurate & comprehensive text measurement & layout

STS2_Editor. C#

Slay_The_Spire_2_AIBot. C#

STS2MCP. Full agentic runs for Slay the Spire 2. A mod that exposes in-game state, and the MCP server for the mod.

shiji-kb. HTML

tooscut. Professional video editing, right in your browser. Made with Rust, WebGPU, WASM, and Tanstack Start.

OpenAlice. Your one-person Wall Street. An AI trading agent covering equities, crypto, commodities, forex, and macro — from research through position entry, ongoing management, to exit.

codex-proxy. codex openAI Responses API => openAI chat Completions API

kimodo. Official implementation of Kimodo, a kinematic motion diffusion model for high-quality human(oid) motion generation.

inspatio-world. Python

Motion_Editor. TypeScript

qmd. mini cli search engine for your docs, knowledge bases, meeting notes, whatever. Tracking current sota approaches while being all local

context-hub. JavaScript

voltagent. AI Agent Engineering Platform built on an Open Source TypeScript AI Agent Framework

openclaw. Your own personal AI assistant. Any OS. Any Platform. The lobster way. 🦞

youtu-rag. Youtu-RAG: Next-Generation Agentic Intelligent Retrieval-Augmented Generation System

autoresearch. AI agents running research on single-GPU nanochat training automatically

video-subtitle-remover. 基于AI的图片/视频硬字幕去除、文本水印去除，无损分辨率生成去字幕、去水印后的图片/视频文件。无需申请第三方API，本地实现。AI-based tool for removing hard-coded subtitles and text-like watermarks from videos or Pictures.

HachiTune. A lightweight pitch editor built with JUCE, featuring neural pitch detection and vocoder resynthesis for real-time preview.

warp-parse. Focusing on building industry-leading ETL engines.

hermes-agent. The agent that grows with you

go-music-dl. 一个基于 Go 语言的全网音乐搜索与下载工具。支持 CLI 命令行与 Web 服务双模式，内置网易云、QQ、酷狗、Bilibili、汽水音乐等 10+ 个主流平台，支持多源并发搜索与无损音质解析。music-dl交流群：1074285005

steerling. Interpretable Causal Diffusion Language Models

agent-browser. Browser automation CLI for AI agents

nullclaw. Fastest, smallest, and fully autonomous AI assistant infrastructure written in Zig

SteptronOss. A lightweight, AI-native training framework for large language models. Designed for fast iteration, reproducible experiments, and modular configuration across SFT, RLVR, and evaluation workflows.

Surge. Blazing fast TUI download manager built in Go for power users

magpie. Multi-AI adversarial PR review tool

moyin-creator. AI 影视生产级工具 | 支持 Seedance 2.0 | 剧本到成片全流程批量化 | AI-powered film production tool with Seedance 2.0 support

spark. :sparkles: An advanced 3D Gaussian Splatting renderer for THREE.js

voxtral-mini-realtime-rs. Voxtral ASR & TTS running natively and in the browser. A Rust implementation of Mistral's Voxtral mini realtime ASR / TTS using the Burn ML framework

Qwen3-ASR-GGUF. 将 Qwen3-ASR 的 LLM 部分导出为 GGUF，用 llama.cpp 进行加速推理。后者支持 Vulkan 和 Cuda 加速。

Mobile-O. [CVPR'26 Demo] Mobile-O: Unified Multimodal Understanding and Generation on Mobile Device

CodeFuse-Embeddings. Text and code embeddings research from CodeFuse: C2LLM, D2LLM, E2LLM, F2LLM, ML-Embed

zeroai. learn AI from scratch

Folio-OCR. Open-source batch OCR workbench — a free, local alternative to ABBYY FineReader. Powered by Ollama + GLM-OCR + PP-DocLayoutV3, ~0.5s/page on RTX 4090. Three-panel editor, layout-aware, PDF/image batch processing, Markdown/Word export. 批量OCR工作台，纯本地运行，免费平替ABBYY，适合书籍文档数字化。

Bittorrent-Client-in-Go. A distributed P2P BitTorrent client in Go with custom networking and concurrency handling.

zvec. A lightweight, lightning-fast, in-process vector database

pi_agent_rust. High-performance AI coding agent CLI written in Rust with zero unsafe code

franken_engine. Native Rust runtime for adversarial extension workloads with deterministic replay, cryptographic decision receipts, and fleet-scale containment.

cli. 📜 Entire CLI hooks into your Git workflow to capture AI agent sessions as you work. Sessions are indexed alongside commits, creating a searchable record of how code was written in your repo.

tambo. Generative UI SDK for React

bunqueue. ⚡ High-performance job queue for Bun. SQLite persistence, DLQ, cron jobs, S3 backups. Built for AI agents and automation

sql-tap. Watch SQL traffic in real-time with a TUI

monty. A minimal, secure Python interpreter written in Rust for use by AI

CommonGround. From isolated agents to shared work

hapi. App for Claude Code / Codex / Gemini / OpenCode, vibe coding anytime, anywhere

zureshot. Pixel-perfect screen recorder for macOS. HEVC + GIF. Zero-copy GPU pipeline. 3.9 MB.

claude-task-viewer. A web-based Kanban board for viewing Claude Code tasks

nanocode. Minimal Claude Code alternative. Single Python file, zero dependencies, ~250 lines.

Mantic.sh. A structural code search engine for Al agents.

frankentui. Minimal, high-performance terminal UI kernel with diff-based rendering, inline mode, and RAII terminal cleanup

BashClaw. Bash is All You Need. A pure Bash reimplementation of OpenClaw. No dependencies. No runtime. Runs everywhere since 2006

voxtral.c. Pure C inference of Mistral Voxtral Realtime 4B speech to text model

vscode-code-kingdom. VSCode插件：像游戏领地图一样，根据git一键显示整个项目中不同开发者之间的势力范围图

codag-visualizer. Visualize AI/LLM workflows in your codebase.

Maestro. Agent Orchestration Command Center

vibe. Easy Linux virtual machine on MacOS to sandbox LLM agents.

mux. A desktop app for isolated, parallel agentic development

presenton. Open-Source AI Presentation Generator and API (Gamma, Canva, Beautiful AI, Decktopus, Presentations AI Alternative)

beautiful-mermaid. TypeScript

apptron. Local-first development platform

pigsty. Enterprise-Grade OSS PostgreSQL Distribution with HA, PITR, IaC, Monitor, 12 kernel forks and 500+ PG extensions. Best-of-breed products integrated as a platform. Self-host Postgres like a Pro!

pad.ws. Whiteboard as an IDE, draw and code in your browser

163MusicLyrics. 云音乐歌词获取处理工具【网易云、QQ音乐】

two-moons. 🌙 Powerful and Interactive Music Learning Platform.

astra. ASTRA is an end-to-end system for synthesizing agentic trajectories and rule-verifiable environments for SFT and RL training, developed by Beike Language and Intelligence (BLI).