Efficiency toolsPerplexicaPWAI search, Perplexity alternatives, open source, multimodal
Perplexica
An open-source AI search engine, an alternative to Perplexity, supporting multiple search modes
5871
View repositoryIndependent Column
Open-source projects worth exploring
Runnable, documented projects worth studying and extending.
Efficiency toolsPerplexicaPWEfficiency toolsWindsurfPWCode editor, AI-first, intelligent coding, Codeium
Windsurf
AI-first code editor, deeply integrated with AI for intelligent coding and project understanding
5876
View repository
Efficiency toolsA tool for real-time multi-AI collaboration: Claude Code Bridge, allowing Claude, Codex, Gemini, and OpenCode to work side by side on the same interface, with visibility, control, and continuity throughoutPWCLI, AI, open source
A tool for real-time multi-AI collaboration: Claude Code Bridge, allowing Claude, Codex, Gemini, and OpenCode to work side by side on the same interface, with visibility, control, and continuity throughout
Claude Code Bridge is a real-time multi-AI collaboration tool that supports Claude, Codex, Gemini, and OpenCode working in parallel within the same interface. Each model can locally persist conversation history and supports one-click recovery. This tool uses a pure CLI workflow with automatic layout, suitable for tasks such as code review, architecture design, and troubleshooting
5876
View repository
OthersYiPWZero-One Everything, large models, multimodal, open source
Yi
Zero One Universal Open Source Large Language Model Series supports multiple specifications and multimodal modes
5869
View repositoryEfficiency toolsTencent Hunyuan's 3D generative model has just released its latest version: v2.5, with the total parameter size increased from 1B to 10B, and the number of effective dough pieces increased more than tenfoldPW3D generation, modeling
Tencent Hunyuan's 3D generative model has just released its latest version: v2.5, with the total parameter size increased from 1B to 10B, and the number of effective dough pieces increased more than tenfold
Tencent Hunyuan 3D Generation Model v2.5 has been released, increasing the parameter count from 1B to 10B, increasing the number of effective patches tenfold, supporting multi-view PBR veneers, 4K high-definition textures, and fine-grained bump mapping, optimizing the bone skinning system and improving 3D animation generation efficiency
5872
View repository
OthersDifyPWLLM platform, RAG, low-code, workflow
Dify
An open-source LLM application development platform offering visual prompt orchestration, RAG pipelines, agent frameworks, and model management
5876
View repository
Efficiency toolsOpenRouterPWModel routing, API, multi-model, optimization
OpenRouter
AI model routing platform: a single API accesses hundreds of AI models and automatically selects the optimal model
5871
View repositoryOthersBettaFish WeiyuPWAI, Agent, public opinion analysis, Python
BettaFish Weiyu
A multi-agent public opinion analysis assistant accessible to everyone, breaking the information cocoon, restoring public opinion to its original state, and predicting future trends. Built from scratch, no framework reliant, starred 38k+.
5869
View repositoryEfficiency toolsA Claude Code plugin: CallMe, which can proactively call Claude when completing tasks, getting stuck, or needing to make decisionsPWAI phone assistant
A Claude Code plugin: CallMe, which can proactively call Claude when completing tasks, getting stuck, or needing to make decisions
CallMe is a Claude Code plugin that allows Claude to make calls proactively when completing tasks or making decisions, so users don't have to stare at the computer screen. It supports multi-turn voice conversations, allowing calls to search, run code, and read and write files, and is compatible with smartphones, smartwatches, and landlines
5876
View repository
Efficiency toolsHKUDS has launched a hugely popular AI personalized learning tool: DeepTutor. It solves three problems: too much information to find answers, no one to explain or understand complex knowledge, and no one to set questionsPWAI learning, AI
HKUDS has launched a hugely popular AI personalized learning tool: DeepTutor. It solves three problems: too much information to find answers, no one to explain or understand complex knowledge, and no one to set questions
DeepTutor is an AI personalized learning tool that can solve problems such as difficulty in searching for information, insufficient understanding of complex knowledge, and lack of practice questions. Through interactive visual learning, it breaks down complex concepts into intuitive diagrams and generates exercises or mock papers based on users' knowledge levels and learning goals, greatly improving learning efficiency
5875
View repositoryOthersOpenVikingPWAgent, context management, ByteDance
OpenViking
ByteDance Volcano Engine uses an open-source AI Agent context database that manages agent memory, resources, and skills uniformly through a file system paradigm
5872
View repositoryOthersPromptfooPWPrompt testing, red team testing, AI safety, open source
Promptfoo
Prompt/Agent/RAG testing framework, supporting red team testing and security scanning, comparable to models like GPT, Claude, Gemini, and supporting CLI and CI/CD integration.
5871
View repositoryOthersMobileAgentPWagent, mobile, GUI, multimodal
MobileAgent
Alibaba's X-PLUG open-source mobile GUI Agent series supports multimodal understanding of phone screens and automatic operations, making it a benchmark project for GUI automation
5877
View repositoryEfficiency toolsA "Visio" plugin for Claude Code: cc-wf-studio, which drags and drags zero code to create automated workflows for Claude CodePWNo-code platform
A "Visio" plugin for Claude Code: cc-wf-studio, which drags and drags zero code to create automated workflows for Claude Code
cc-wf-studio is a no-code plugin designed for Claude Code, supporting complex workflow creation, including branches, loops, and user interactions. Users can design workflows through a graphical interface and export them as .claude format files, with built-in AI optimization features that automatically adjust nodes and routes
5876
View repositoryAI and Agentsclaude-skillsPWOpen source projects
claude-skills
A collection of 169 production-ready Claude Code / OpenAI Codex / OpenClaw skills and plugins covering engineering, marketing, product, compliance, and more.
5876
View repositoryOthersShanghai AI Lab has open-sourced InternVLA-A1, an end-to-end unified model covering vision, language, and actionPWVLA, open source, model
Shanghai AI Lab has open-sourced InternVLA-A1, an end-to-end unified model covering vision, language, and action
The Shanghai AI Lab has open-sourced InternVLA-A1, a unified visual-language-action model that can perform tasks under natural language prompts, such as "putting a pen on the desk into the pen holder" without presetting coordinates, solving the problem of information loss between traditional perception and action, and supporting multimodal perception and understanding
5875
View repositoryEfficiency toolsAiderPWTerminals, AI programming, Git, code editing
Aider
AI programming assistant in the terminal allows direct editing of code files through conversation, supporting Git integration
5775
View repositoryEfficiency toolsMiroFish - Swarm Intelligence Prediction EnginePWAI, prediction, swarm intelligence, Python
MiroFish - Swarm Intelligence Prediction Engine
A simple and universal swarm intelligence engine that can predict everything. Based on Swarm Intelligence theory, implemented in Python, it supports various prediction scenarios. Today, GitHub has 4400+ stars.
5775
View repositoryEfficiency toolsAn AI programming agent task board tool: Vibe Kanban, which helps you amplify AI programming efficiency by 10 timesPWAI programming, AI
An AI programming agent task board tool: Vibe Kanban, which helps you amplify AI programming efficiency by 10 times
Vibe Kanban is an AI programming agent task board tool that can significantly improve programming efficiency. It supports task management for multiple agents, real-time monitoring of status and results, automatically creating git branches and committing changes, making it convenient for developers to review code and perform pull requests. Users can track task execution through centrally managed MCP configurations and detailed logs
5776
View repository
OthersGLM-4PWZhipu, bilingual, 128K, multimodal
GLM-4
Zhipu AI's open-source bilingual dialogue model supports 128K context and multimodal understanding
5773
View repositoryOthersMiroFishPWMulti-agent, prediction, swarm intelligence, simulation
MiroFish
A multi-agent swarm intelligence prediction engine simulates future trends by building a high-fidelity digital world. Upload seed materials and describe forecast needs in natural language to obtain detailed forecast reports and interactive digital sandboxes.
5775
View repositoryOthersMuseVPWVirtual humans, video generation, diffusion models, Tencent
MuseV
A virtual human video generation framework based on diffusion models, supporting unlimited video lengths
5773
View repositoryEfficiency toolsClaude CodePWTerminal, AI programming, Anthropic, command line
Claude Code
Anthropic's terminal AI programming tool collaborates with Claude directly on the command line
5776
View repositoryOtherspromptfooPWPrompt testing, red teaming, evaluation, safety
promptfoo
LLM application evaluation and red team testing tools support automated prompt testing, model comparison, security vulnerability scanning, and CI/CD integration, compatible with mainstream model providers such as OpenAI, Anthropic, and Ollama.
5775
View repository