GitHub repos

A workflow-automation platform that combines a visual canvas with custom code and native AI nodes — 400+ integrations to wire agents into real systems.

Pull and run Llama, DeepSeek, Mistral, Qwen, Gemma, and dozens of other open models locally with a single command and a clean API.

A huge community collection of "act as …" prompts (now prompts.chat). Paste a role, get a working persona for ChatGPT, Claude, or any chat model.

The de-facto model-definition framework: load, run, and fine-tune state-of-the-art transformers for text, vision, and audio with a unified API.

A visual builder for AI agents and retrieval pipelines — wire up models, tools, and logic on a canvas, then export or serve it as an API.

A production-ready platform for building agentic apps: a visual workflow builder, RAG pipelines, model management, and an API to ship from.

A polished, self-hosted web interface for Ollama and any OpenAI-compatible API, with chat history, RAG, and multi-user support.

A modular framework for building LLM apps — composable chains, agents, memory, and integrations with hundreds of tools and vector stores.

An agentic coding tool that lives in your terminal, understands your whole codebase, and edits files, runs tests, and opens PRs from plain English.

High-performance LLM inference in plain C/C++ with quantization — the runtime under much of the local-model ecosystem, from laptops to Raspberry Pis.

The most powerful and modular diffusion GUI — a node graph for Stable Diffusion, Flux, and video models, with full control over every step.

21 hands-on lessons that take you from "what is a token" to shipping generative-AI apps, with code in Python and TypeScript.

An open-source command-line agent that brings Gemini’s multimodal model into your terminal for coding, automation, and everyday tasks.

Make websites accessible to AI agents — drive a real browser to click, type, and extract, so agents can automate web tasks end to end.

The official collection of Model Context Protocol servers — filesystem, git, fetch, and more — that give Claude, Cursor, and other clients real tools.

An open-source RAG engine built on deep document understanding — chunking, citations, and agentic retrieval for enterprise knowledge bases.

A fast, memory-efficient inference and serving engine using PagedAttention — the production server behind many open-model deployments.

An AI developer (formerly OpenDevin) that writes code, runs commands, browses the web, and fixes its own mistakes inside a sandbox.

Guides, papers, lectures, and notebooks covering prompt and context engineering techniques — few-shot, chain-of-thought, ReAct, and beyond.

Official examples and guides for common tasks with the OpenAI API — embeddings, function calling, evals, and retrieval.

A programming framework for agentic AI where multiple agents collaborate, call tools, and hand off tasks to solve problems together.

A framework for orchestrating autonomous agents with defined roles, goals, and tools that collaborate as a "crew" on a shared task.

A data framework for LLM apps — ingest, index, and query documents, then build retrieval and document-agent workflows on top.

A collection of notebooks showing effective patterns with Claude — tool use, sub-agents, retrieval, vision, and prompt-caching recipes.

An open framework for augmenting humans with AI: a library of reusable prompt patterns (summarize, extract wisdom, write essays) you can pipe into from the terminal.