AI That Decides What Matters
Open-source AI tools, local inference engines, and research platforms — built for developers, researchers, and teams who want full control over their AI stack.

WorthDoing AI is an open-source AI organization focused on building tools that empower developers, researchers, and teams to harness the full potential of artificial intelligence — without vendor lock-in, without black boxes, and without compromises on privacy or performance.

We believe in:

A unified AI playground to explore, test, compare, and understand AI models — powered by OpenRouter with 350+ models across 55+ providers in one interface.

A production-grade, open-source application that serves as a single interface to explore, test, compare, and understand AI models from all major providers through the OpenRouter unified API. Built by WorthDoing AI.

Key Features:

Tech Highlights:

• Server-side API key protection via secure proxy layer
• Dark-mode first design with responsive layout
• Real-time streaming with Server-Sent Events (SSE)
• Advanced model filtering by capability, context length, pricing, and provider
• Full parameter control: temperature, top-p, top-k, frequency penalty, max tokens
• Conversation export and history management

A custom C++ inference engine for Apple Silicon — run large language models locally with maximum performance, complete privacy, and zero cloud dependencies.

This is not a wrapper around llama.cpp or any existing runtime. It is a home-made C++ inference engine built from scratch, capable of loading and running GGUF quantized models, exposed through a polished local web application with direct HuggingFace model downloading. Built by WorthDoing AI.

What Makes This Different:

Engine Architecture:

Web UI Features:

• Real-time chat with streaming token display
• Model library with file-level management
• Built-in HuggingFace browser for model discovery and one-click downloads
• Performance monitoring dashboard with tokens/sec, memory usage, and system metrics
• System settings and configuration panel

Next-gen Mac system monitor — real-time CPU, RAM, GPU, disk, network, and battery in a beautiful interactive terminal UI.

npm install -g macpulse
macpulse

4 view modes: Dashboard, Detailed (d), Process (p), Graph (g). Gradient progress bars, rounded-corner panels, dual-column layout on wide terminals.

A local AI agent operating system for your terminal — think, act, execute, persist.

npm install -g wd-agent
wdagent

Uses Anthropic's native tool_use API (zero JSON parsing), WorthDoing Capabilities as the execution layer, context auto-compaction, and token tracking. Supports both Anthropic direct and OpenRouter (350+ models) with interactive scrollable model selection.

Production-grade capability package — 31 portable, composable agent actions across 8 API providers.

Python package: runtime, registry, YAML contracts, CLI (wdcap), 188 tests. TypeScript SDK: zero dependencies, ESM + CJS, interactive CLI explorer.

Official logo and brand assets for WorthDoing AI — 12 variants in SVG and PNG formats.

The WorthDoing AI icon represents a neural network converging toward a single decision core — a visual metaphor for AI that filters, evaluates, and selects high-value actions. The gradient progression blue → cyan → green represents the transformation from understanding to decision.

Available Variants:

Repository	Description	Language	Status
termxl	Terminal-native spreadsheet — Excel in your terminal with formula engine (npx termxl)
macpulse	Next-gen Mac system monitor — CPU, RAM, GPU, disk, network, battery (npx macpulse)
wd-agent	Local AI agent OS — Claude Opus 4.6 + WorthDoing Capabilities (npx wd-agent)
ai-playground	Unified AI playground — 350+ models, compare lab, embeddings, image generation
local-llm-inference	Custom C++ inference engine for Apple Silicon with professional Web UI
worthdoing-capabilities	Official capability package — portable, composable agent actions with runtime, registry, and CLI
worthdoing-capabilities-js	TypeScript SDK — same capabilities on npm (worthdoing-capabilities)
worthdoing-website	Official website — landing page + project showcase
brand-assets	Official logos and brand assets — 12 variants in SVG and PNG

Phase	Milestone	Status
Q2 2026	Metal GPU acceleration for Local AI inference engine
Q2 2026	Custom .wdf model format with optimized loading
Q2 2026	AI Playground — multi-turn conversation branching
Q3 2026	Fine-tuning toolkit for local models
Q3 2026	Model benchmarking suite with automated evaluation
Q3 2026	Plugin architecture for community extensions
Q4 2026	Cross-platform Local AI support (Linux, Windows)
Q4 2026	Federated model sharing across local instances
Q4 2026	Enterprise deployment toolkit

# System monitor — see your Mac in real time
npm install -g macpulse
macpulse

# AI Agent — local agent OS powered by Claude
npm install -g wd-agent
wdagent

# Capabilities SDK — 31 portable agent actions
npm install worthdoing-capabilities

# AI Playground
git clone https://github.com/Worth-Doing/ai-playground.git
cd ai-playground && pnpm install && pnpm dev

# Local LLM Inference
git clone https://github.com/Worth-Doing/local-llm-inference.git
cd local-llm-inference/engine && make -j$(sysctl -n hw.ncpu)
cd ../app && pnpm install && pnpm dev

# Website
git clone https://github.com/Worth-Doing/worthdoing-website.git
cd worthdoing-website && npm install && npm run dev

We welcome contributions from the community! Whether you're fixing a bug, improving documentation, or proposing a new feature, we'd love to hear from you.

How to contribute:

1. Fork the repository you want to contribute to
2. Create a feature branch (git checkout -b feature/amazing-feature)
3. Commit your changes (git commit -m 'Add amazing feature')
4. Push to the branch (git push origin feature/amazing-feature)
5. Open a Pull Request

Contribution areas we're especially interested in:

All code repositories are released under the MIT License — free to use, modify, and distribute.

Brand assets (logos, icons, visual identity) are proprietary to WorthDoing AI and may not be used without permission.

Built with purpose by WorthDoing AI
AI That Decides What Matters