Trending Repos

AI coding agent designed for small language models, achieving 87% benchmark with a 4B-active model.

Why: Enables efficient AI coding for developers using smaller language models.

A library of HTML slide templates designed so any coding agent can pick the right one and produce a beautiful deck on the user's behalf, automatically.

TokenSpeed is an inference engine designed to accelerate large language model (LLM) computations.

Why: It enables faster processing of LLMs, which is crucial for real-time applications and improving overall performance.

ChatGPT Team(Business) 促销码自动扫描工具 — 批量发现/验证/价格收集，支持 17 国 34 个码，最高折扣 71% | ChatGPT Business promo code scanner — batch discovery, validation, price collection, 34 codes across 17 countries, up to 71% off

🔍 The hardest search benchmark in the wild — vague, multi-turn, proactive. 200 long-horizon tasks with persona-driven progressive disclosure, scored by verifiable schema-free knowledge-graph evaluation. No vibes, just triplet F1.

Personal-Model First Self Evolving AI Agent 🐘

KVarN is a native vLLM KV-cache quantization backend for your agents: 3-5x more context, throughput above FP16, and FP16-level accuracy. Calibration-free, one flag.

Fast, lossless LLM inference via dual-view diffusion decoding.

An advanced, LLM-powered toolkit providing comprehensive capabilities for ethical synthetic media detection, analysis, and responsible content generation.

Surreal dream sequence engine — procedural narrative generation from emotional state, memory anchors, and symbolic archetypes

🔮 GPT-Image-2 Mastery 2026: Ultimate AI Prompts & Editing Toolkit

让 AI 住在你 MacBook 的刘海里 · 零依赖开箱即用 · 多引擎并行的桌面 AI 伴侣（Swift 6 / SwiftUI / macOS 14+）

An open-source Cotypist with macOS system wide AI autocomplete

Security-native LLM system for AI-generated application security.

Surreal dream sequence engine — procedural narrative generation from emotional state, memory anchors, and symbolic archetypes

End-to-end AI short-video production pipeline. FastAPI orchestration + Spring Boot gateway with multi-model failover, circuit breaker, metering, and full-stack observability. AI quality gating: prompt anchoring, CLIP consistency, AV sync auto-rescue.

An end-to-end framework for multi-speaker transcription that jointly models who spoke, when, and what.

A recursive evidence-gated cognitive runtime for memory-native AI agents, combining hybrid retrieval, temporal reasoning, async learning, and plug-and-play tools.

Inference-native Tokenmaxxing Agent Harness for Loop Engineering

Estimate whether a Hugging Face model fits and fine-tunes on your local GPU.

LLMInternSkill: LLM internship resume and job-search Codex Skill for resume polish, JD tailoring, evidence guard, interview grilling, and Project Scout. 大模型实习简历与求职工具箱。

An advanced, LLM-powered toolkit providing comprehensive capabilities for ethical synthetic media detection, analysis, and responsible content generation.

codex app - AI coding tool for writing, reviewing, and debugging code. Comprehensive guide with setup instructions, feature highlights, and system requirements.

Built by DeepSeek, for DeepSeek — a Swift-native macOS coding agent

LLM internship resume and job-search Codex Skill: resume polish, JD tailoring, evidence guard, interview grilling, and Project Scout for LLM/RAG/Agent roles. 大模型实习简历与求职工具箱。

DeepSeek-first desktop coding agent workbench

A hand-curated, topic-organized library of the best ML education — 923 docs (391 arXiv papers, 474 Stanford/MIT/Karpathy/fast.ai lectures, 58 explainer articles), normalized to Markdown with full provenance. Open it in Obsidian or point your agent at it. A clean ML corpus for learning, RAG & fine-tuning.

Share your terminal with coding agent in one command. Lightweight shell proxy, written in Rust.

Token-saving companion for OpenCode — 42 compression layers, zero risk, no caveman speak

The definitive OpenAI, Claude, MCP, Harness, Evals, and Production Agent Systems learning roadmap.

A minimal Rust gateway built for coding agents (LiteLLM-Compatible)

Python port of Claude Code's agent-runtime architecture, on LangChain

Ask several AI models one question — they answer in parallel, critique each other blind, and you decide. Native macOS, bring-your-own-keys.

Code Modern. Code Legacy. Code Firmware. - open-source AI-native IDE with agentic coding, Power Mode, legacy modernization, and firmware development

LLMs have ended reverse engineering as a high-barrier skill. Case study: reconstructing 6/7 custom HTTP signing parameters of Douyin (TikTok CN) v38.1.0 native library in 30 days using Claude Opus. 大型语言模型（LLMs）已经终结了“逆向工程是一项高门槛技能”的时代。案例研究：使用 Claude Opus，在 30 天内成功还原抖音（中国版 TikTok）v38.1.0 原生库中 7 个自定义 HTTP 签名参数中的 6 个。

Turn Claude Code into its own Meta-Harness — a skill that evolves the scaffolding around a fixed model (memory, retrieval, context, prompts) via a native propose→score→Pareto loop. Native reimplementation of Meta-Harness (Lee et al. 2026).

Codex-first conversion of Karpathy-inspired coding agent guidelines.

Don't make LLMs honest. Make every factual claim auditable. — An LLM Claim Auditing Layer with T1-T7 truth gradients. 98.1% business effectiveness on LiarBench v0.2.

Whisper Desktop Pro - OpenAI Whisper-based desktop transcription for offline use. Includes setup guide, feature overview, and download instructions.

MemPrivacy is a privacy-preserving personalized memory management framework for edge-cloud agents.

Experimental open-source macOS menubar app for speech-to-text workflows

AgentGuard：An Attribute-Based Access Control Framework for Tool-Use LLM-Based Agent

AI 文献检索与综述 Agent：支持多源检索、代码仓库定位、开放 PDF 下载、证据链与中英双语报告。

Hermes Agent plugin for cron observability. Turn hidden automation into visible spend.

玄霖超脑 · 无量网络 v4 重构版— 66 模块6通道 AI 认知神经系统 · 装一次，你所有的 AI 工具从此共享一个永远不失忆的大脑。跨会话记忆 · 多 Agent 互通 · 自动代码审查 · 零配置

Editable, part-aware 3D generation from text or reference images. Open-source client for nova3d.xyz.

Dataflow-Oriented Reinforcement Learning for (Multi-)Agentic LLMs

使用 QQ/微信 聊天记录结合向量数据库让AI更好的扮演对方的角色，在不微调模型的情况下可以达到可观的效果。把曾经的美好，续成往后的陪伴。 --续温

Official Implementation of LongLive-RAG: A general retrieval-augmented framework for long video generation.

Implementation of D4RT, Efficiently Reconstructing Dynamic Scenes, from Deepmind

Open-source creative asset workspace for teams and AI agents 🤖

Open Agent OS for Claude Code, Codex, and Cursor: meta-agent builder, A2A Hub routing, local ontology, memory and security gates.

Turn raw footage into brand-ready, platform-optimized video with one command. Local-first: FFmpeg + WhisperX/MLX + Ollama.

A survey of rubrics across the evolving LLM landscape.

《如何从零构建 7×24 小时 AI Agent》—— 拆解 30 万行真实 AI Agent 平台的技术书

🩵uncensored Fully offline AI girlfriends harem Openclaw+Local LLM+GPT-SoVITS+ComfyUI image+Live2D+desktop pet+SilllyTavern Character card import| Dual channels for QQ & Telegram | Dynamic 8GB VRAM scheduling, can run offline完全离线 AI 女友后宫 |小龙虾+本地大模型+GPT-SoVITS+ComfyUI+Live2D+桌宠+酒馆角色卡导入|QQ & Telegram 双通道 | 8GB 显存动态调度断网可跑

📬 Drop-in feedback widget for websites & apps. Turn submissions into structured email & webhook data for AI agents. Framework-agnostic Web Component. Accessible, customisable, zero deps.

🤖 Resources for AI product managers.

Personal usage analytics for Pi coding agent. Temporal-aware insights, context-aware suggestions, model efficiency analysis.

A curated list of AI-powered web scraping tools, LLM-friendly crawlers, MCP servers, and infrastructure for turning the web into data.

Implementation of Kalmanformer, modeling the Kalman gain with a transformer

Alephant AI Gateway is an open source, OpenAI-compatible AI Gateway for 50+ providers, 320+ models, and custom model backends.

From-scratch OTP-native Elixir implementation of LangChain, LangGraph, and Deep Agents concepts

A terminal coding agent for real project work.

一个面向架空世界故事创作者的 AI 架空世界与长篇故事引擎。它能从一句设定出发，自动生成层 层嵌套的历史、地理、角色关系与事件网络，并持续写出贴合世界观的剧情。支持 CLI 与可视化界面，可按时间和空间路径扩展世界，让故事像真实宇宙一样不断生长。项目用数字路径管理嵌套的时间、空间、 角色和事件节点，按需生成并补全世界资料，支持本地 Demo、OpenAI/DeepSeek JSON Agent、CLI 和黑白 Web 可视化界面，用于创建可持续展开的故事世界与剧情会话。

A personal AI agent that can work safely on your machine, remember useful context, and keep its data under your control.

training your 0.1B llm from zero

ChatGPT — gpt 5. for Windows

👻 AI-Powered Pentesting Command Center — Autonomous security testing with real-time streaming, self-improving AI, unlimited tool iterations, and beautiful dark UI

Drop-in skill so coding agents (Claude Code / Codex) can search the Banodoco Discord message feed for video/image generation best practices.

OpenAI Assistants API Setup — openai assistant setup, gpt assistant tutorial, openai api agent, build ai assistant.

AI Reverse Proxy Collection — Collection of AI reverse proxies for free access to GPT-5, Claude Opus, and other models for Windows

Lightweight agent multiplexer, all in one Web dashboard

Self-evolving context for any coding agent.

MemTrace: Tracing and Attributing Errors in Large Language Model Memory Systems

Premium CLI tool for bulk ChatGPT account creation via direct HTTP fetch. No browser needed. Parallel processing, live progress UI, auto OTP verification.

Whisper Desktop Pro - OpenAI Whisper-based desktop transcription for offline use. Includes setup guide, feature overview, and download instructions.

codex app - AI coding tool for writing, reviewing, and debugging code. Comprehensive guide with setup instructions, feature highlights, and system requirements.

Fix "Codex cannot access its local database" / "migration 1 was previously applied but has been modified" / "timed out waiting for state db backfill after 30s" on Codex Desktop (Windows/WSL/Mac/Linux) after the 0.130 → 0.131 update. CRLF/LF sqlx checksum drift in state_5.sqlite + logs_2.sqlite. Sessions preserved. See openai/codex#23787.

GitHub Action + CLI that auto-writes PR descriptions, adds labels, and reviews code using OpenAI

Chrome extension wallet for user-owned AI API access

coding agent

Retrieval is CheapShow Me the Code: Executable Multi-Hop Reasoning for Retrieval-Augmented Generation

An open-source End-to-End AI model for self driving

A step by step implementation to build an agent that runs for week to complete a task and self improve.

Implementation of WorldParticle, Unified World Simulation of Lagrangian Particle Dynamics via Transformer

Implementation and explorations into DiscoRL, Discovering state-of-the-art reinforcement learning algorithms, David Silver's last work at Deepmind

A visual, editable context layer for Codex. AI edit AI's context with surgical precision, giving you more control and freedom over what Codex sees.

Open Source Agentic Security Scanner for your codebase.

该项目是无人机的租赁系统，也可以是其他的设备、器材租赁，技术栈：Java 21 + Spring Boot 3.5.13 + Vue 3 + Spring AI + MCP + AI Memory，其他的技术栈Dify/Coze、 Python LangChain/LangGraph Agent 。项目是开源的，免费学习的，star支持一下，不断更新中。

Implementation of RigidFormer, Learning Rigid Dynamics using Transformers

ChatGPT Plus Desktop — OpenAI ChatGPT Plus Desktop 2026 — GPT-5 access, image generation, voice, custom GPTs. ChatGPT Plus Desktop premium subscription unlocked free, full account access, all features, lifetime activation 2026. No trial limit, license key included, Windows 10/11.

LLM-powered Chrome extension for automatic web form filling. All data stays local.

An open-source End-to-End AI model for self driving

Curated GPT-Image-2 prompt library with preview images, 1-click copy, and direct generate links. Supports multi-languages (EN/ZH/PL/VI/ID/TR). Filter by category (Poster, Portrait, UI/UX, Architecture...). Traceable to original social media posts. Free & daily updated.

Progressive agentic reading harness: sentence maps, bounded chunks, Chronicle Markdown books, packets, validation. Python CLI + Next.js demo. MIT.

OpenAI Sora — full premium build for Windows with pro features unlocked. Creative Suite Pro Pack

Implementation of Attention Schema-based Attention Control (ASAC), A Cognitive-Inspired Approach for Attention Management in Transformers

NeuroLens is a premium, state-of-the-art Document Retrieval-Augmented Generation (RAG) system. It combines an immersive, holographic, sci-fi-themed React frontend with a high-performance FastAPI backend.

Advanced RAG (Retrieval-Augmented Generation) system for private document intelligence. Uses OpenAI GPT-4o and Vector Databases to provide zero-hallucination chat insights.

A curated directory of every first-party skill, capability, and tool across Claude, ChatGPT, Gemini, Codex & DeepSeek — searchable, filterable, deep-linkable.

Desktop realtime speech translator for routing translated audio between virtual audio devices in calls.

The same AI agent pipeline built in Mastra and LangChain. Runs in parallel, measures everything.

Biologically inspired language model using Jaccard Surprise as its only training signal. No backprop. No GPU. Online Hebbian learning from corrections. Two-layer cortex with apical feedback. Runs on CPU under 200MB.

Move beyond prompts. Build verifiable, intent-driven Semantic Agents.

artificial-intelligence computer-vision deep-learning emotion-detection facial-expression-recognition tensorflow opencv cnn machine-learning python real-time-ai

ChatGPT Plus/Pro free trial checkout link generator with auto/manual proxy support

GitHub Action that runs trustabl — static reliability/safety analyzer for AI agent SDKs (Claude, OpenAI, Google ADK, MCP). Gates CI on risk + severity.

Open-source AI Operating System (AIOS) blueprint for Claude Code & Codex — turn any agent into your personal second brain + Chief of Staff: skills, self-verifying workflows, a second-brain wiki, and deterministic tools. Fork it for any role or domain.

Free-tier AI load balancer — auto-rotates 28+ keys across 9 providers (Gemini, Groq, Mistral & more). OpenAI-compatible drop-in. Works with any agent framework.

An elegant LLM chatbox for myself that runs in the browser.

Talk to your Mac — a native macOS voice assistant that sees your screen, drives your Mac, and answers back. Powered by the OpenAI Realtime API. Open source.

一个生产级的深度研究 Agent 系统，从零构建多智能体编排、Red-Blue 对抗降噪、 语义级上下文压缩、跨 Agent 共享记忆四大核心能力，配套 165 次独立实验 + Bootstrap 统计显著性检验的完整评测体系。

Sub2API gateway subscription → Codex CLI / Codex App injector. macOS-native, zero-deps Python single-file. Supports codex2api.com and any self-hosted Sub2API instance.

The complete open-source LLM curriculum — prompt engineering, RAG, fine-tuning, agents, and production deployment. Beginner to advanced. Python. Free forever.

Implementation and explorations into PopuLoRA, Co-Evolving LLM Populations for Reasoning Self-Play

Command-line interface for querying large language models

OpenCode plugin: image generation via your ChatGPT subscription or OpenAI API

AI developer that generates LangChain Expression Language code

Clinical Rag AI

A desktop-first PDF workspace with local-first UX plus retrieval-augmented intelligence. Lumenfolio indexes selected PDF folders, enables accurate document-grounded chat and Q&A, and supports bilingual reading/translation workflows with fast context switching for research and study productivity.

ChatGPT Session ↔ 9 种主流认证格式 N×N 双向互转（auth.json / Cockpit / CPA / Sub2API / 9router / AxonHub / Codex-Manager / Codex Auth / Raw）· 11 种来源自动识别 · Plus 多区域 / Team 工作区订阅长链一站生成（v2.4.0 Stripe init 三步法长链引擎，拿 pay.openai.com 可靠长链）· 油猴 UserScript + Chrome/Edge/Firefox MV3 扩展双发布 · 全程本地处理零上报 · Specimen 设计语言

Lightweight Python SDK for LLM inference logging and observability

The AI-native wire format for structured data. 100% comprehension on every frontier model. 71% fewer tokens than JSON. 33B+ lossless round-trips across 5 formats. Spec v3.1 Stable.

A more personal intelligence, in an interface you already use every day.

Search the web in Pi through your Codex subscription.

Pool the free tiers of 18 LLM providers (Groq, Cerebras, NVIDIA NIM, Gemini, OpenRouter, Cloudflare, Hugging Face…) behind one OpenAI-compatible /v1 endpoint. Free, zero-config, automatic failover. Codex/agent ready.

When Should Models Change Their Minds? Contextual Belief Management in Large Language Models

RedHop — a reasoning-aware retrieval & context runtime for RAG (in-process, no vector DB). Light weight alternative to langchain and llamaindex for retrieval alone.

Codex 保安: local GUI guardrail and monitoring for Codex

FRIDAY is an autonomous cognitive AI operating system for memory, reasoning, voice, cybersecurity, and self-improvement. It runs locally and is designed to move toward general intelligence. Built by a self-taught 17-year-old developer.

Make any web app voice-controllable in 37 languages. Drop-in widget + open-source agent server — visitors speak, the agent talks back and drives their browser in real time.

Forty hands-on recipes for production-grade retrieval-augmented generation, Nebius-first and provider-agnostic.

Docker sandbox backend for deepagents

Pi extension for switching between multiple OpenAI Codex accounts

Your AI's memory grows forever. Your token bill doesn't. A cross-provider memory layer for LLM apps.

GitHub Action that runs trustabl — static reliability/safety analyzer for AI agent SDKs (Claude, OpenAI, Google ADK, MCP). Gates CI on risk + severity.

AGI-powered social network bot for automated interactions

Deterministic, scripted, resumable multi-agent orchestration for LangChain deepagents — a community port of Claude Code's Dynamic Workflows.

Workflow framework for AI agents. MCP-native typed graphs, visual editor, Python export.

OpenGem 💎 - a minimalistic, all-local, fully containerized, open-source AI agent.

Translate PDF documents using LLMs. Preserves original layout.

Sovereign agentic CLI — Textual TUI, multi-provider LLM mesh, persistent sessions, skills, MCP, Telegram gateway, autonomous yolo mode

local-llm ollama llama-cpp vllm deep-research research-agent rag retrieval-augmented-generation agent llm-agent local-first privacy offline qwen searxng faiss bm25 python cli ai-research llm-tooling

Give an AI agent read/write/deepen access to your real Obsidian vault as a living knowledge graph. Zero dependencies. Works with any LLM framework.

A Retrieval-Augmented Generation (RAG) based AI Teaching Assistant that delivers accurate, context-aware answers from custom educational resources using LLMs, embeddings and semantic search.

G-DRAGON: Geospatial Reasoning and Dynamic Planning for Retrieval-Augmented Outdoor Navigation

OpenAI-compatible proxy server for CommandCode API. Exposes /v1/chat/completions and /v1/models endpoints so any OpenAI-compatible client can use CommandCode models through a local HTTP server.

A high-throughput semantic routing engine for AI agents and LLM applications. Executes local intent classification using ONNX embeddings, FAISS indexing, and asynchronous micro-batching.

Sub-second RAG regression testing. Define golden questions, detect lost chunks in CI. pytest for your RAG pipeline.

The cockpit for operators running AI at scale. Local-first Discord-style dashboard for Claude Code sessions across many projects.

An AI-powered coding assistant that helps users generate code, detect errors, explain programming concepts, and improve coding productivity.

Implementation and explorations into MPO / DMPO

OWASP ASI-aligned red-teaming and evaluation framework for AI agents

笃行智元 Agent 应用开发者共学社区：面向 AI Agent、LLM App、工作流自动化与智能体产品开发的学习、实践与项目协作空间。

SageSched: Intelligent LLM Request Scheduler with Workload Prediction — QoS-aware dual-queue scheduling for black-box LLM APIs (OpenAI/Azure/Doubao/Gemini)

A practical, continuously updated learning resource covering every layer of Retrieval-Augmented Generation — from how to split a document to how to evaluate whether your system actually works. Each technique comes with runnable Python code, a Feynman-style explanation, step-by-step algorithm breakdown, Mermaid diagrams and worked examples.

OpenTelemetry-native, local-first observability for AI agents. DuckDB on disk, Next.js viewer on localhost, no SaaS. Adapters for Anthropic, OpenAI, LangGraph, AutoGen, CrewAI, and Claude Code's native OTel emit.

Source code for our paper ''When Knowledge Is Not Free: Cost-Aware Evidence Selection in Retrieval-Augmented Generation''

Open-source PII firewall for LLM apps. Detect, anonymize and rehydrate sensitive data before it reaches OpenAI, Anthropic or any LLM provider.

The .cv open file format - spec, SDKs, viewer, CLI, server middleware

LangGraph — langchain agent, langchain setup, ai agent workflow.

Open-source AI career assistant built with Rust, Tauri, React, and Ollama for resumes, interviews, ATS analysis, and productivity.

Curated, annotated research on reasoning gaps in large language models — temporal reasoning, causal reasoning, and beyond.

本地中文小说创作 Agent (LangChain 1.0 + LangGraph)

An open-source educational platform for learning artificial intelligence through visual neuron building, training, and testing without writing code.