Open AI Stack

A curated directory of 100% free, open-source, and generous freemium AI tools alongside the cutting-edge silicon built to run them locally or on the cloud with affordability.

TL;DevTech // Open Source & Free AI Directory

⚠️ Note: Free software still needs a provider like Claude or OpenAI or a PC with enough resources to run LLMs locally.
🔍
TOOL DESCRIPTION
🤖 General Agent
An open-source, autonomous AI agent that runs locally to automate tasks, manage files, and act as a proactive personal assistant.
Your personal AI super intelligence. Private, simple and extremely powerful.
An open-source, code-first Python toolkit for building, evaluating, and deploying sophisticated AI agents with flexibility and control.
Build and run agents you can see, understand and trust.
The open-source multimodal AI agent stack connecting cutting-edge AI models and agent infrastructure.
An open-source, self-improving autonomous AI agent designed for persistence and memory, capable of learning and creating its own skills over time.
An open-source SuperAgent framework by ByteDance designed to orchestrate sub-agents, memory, sandboxed code execution, and extensible skills for long-horizon tasks.
An open-source, local-first AI coworker that builds a living knowledge graph from your work sources (Gmail, calendar, transcripts) and stores data locally as Markdown files.
An open-source framework for orchestrating role-playing multi-agent teams — define agents with goals and tools, then coordinate them on complex tasks.
An open-source recreation of the Manus general-purpose agent — runs autonomous multi-step tasks using any LLM without needing an invite.
An open-source Python agent framework from the Pydantic team — type-safe, model-agnostic, with built-in dependency injection and streaming support.
A fair-code, self-hostable workflow automation platform with native AI nodes — build agentic workflows connecting 400+ apps and any LLM.
An open-source, self-hostable automation platform with AI agents and MCP support — a no-code alternative to Zapier you can run yourself.
A visual automation platform connecting 3,000+ apps with AI steps — free tier with a monthly operations allowance.
A no-code platform for building AI agents and automations on a visual canvas — free tier with monthly credits.
A human-in-the-loop automation tool with AI steps and a beginner-friendly builder — generous free plan.
🧠 LLM
A curated guide to the top large language models — comparing capabilities, context windows, and the best use cases for each.
Run large language models locally with a single command. Supports Llama, Mistral, Gemma, and dozens more models on Mac, Windows, and Linux.
A self-hosted, feature-rich ChatGPT-style interface for running LLMs locally via Ollama or OpenAI-compatible APIs — with RAG, tools, and user management built in.
A desktop app for discovering, downloading, and running local LLMs with a chat UI and an OpenAI-compatible local server.
An open-source, offline desktop app for running LLMs locally — clean chat UI with model management and an OpenAI-compatible API server.
Run powerful LLMs locally on CPU or GPU with no internet required — includes a desktop chat app and privacy-focused local AI inference.
The foundational open-source inference engine for running quantized LLMs on CPU and GPU — the backbone powering Ollama, LM Studio, and most local runners.
A unified, OpenAI-compatible API that routes to dozens of models from one endpoint — includes a free tier with multiple free models.
A free web playground and API for Google's Gemini models — prototype prompts and ship with a generous daily request quota.
An ultra-fast inference API serving open models (Llama, Mixtral, and more) on custom LPUs — free tier with thousands of requests per day.
Mistral's developer platform and API for its open and commercial models — free tier with a large monthly token allowance.
An inference API delivering some of the fastest token speeds available for open models — free tier with millions of tokens per day.
NVIDIA's hosted API for hundreds of open models with an OpenAI-compatible endpoint — free credits to get started.
Run open models on Cloudflare's global edge network from a serverless API — generous free daily allocation.
🔌 MCP
The official open-source debugging tool for MCP servers — inspect requests, responses, and tool schemas interactively in a browser UI.
A high-level Python framework for building MCP servers quickly — define tools, resources, and prompts with minimal boilerplate.
A community registry for discovering, installing, and sharing MCP servers — browse thousands of integrations for agents and AI tools.
🔧 Utilities
A browser-based tool that automatically detects your hardware and estimates which AI models can run locally on your system.
An open-source, all-in-one desktop app for chatting with your documents using any LLM — supports RAG, agents, and multi-user workspaces locally.
A free, open-source drop-in replacement for the OpenAI API that runs locally — supports text, image, audio, and embedding models with no cloud dependency.
An open-source alternative to Google NotebookLM — upload documents and generate podcast-style audio summaries using local or open-source TTS models.
An open-source, Python-native vector database for building LLM apps — store embeddings and power search and retrieval (RAG) locally.
An open-source AI browser automation framework built on Playwright — control the web with natural-language actions.
An open-source long-term memory layer for LLMs — give your agents persistent recall across sessions for more coherent conversations.
Google's AI research notebook that grounds answers in your sources and generates audio overviews — free to use.
An AI meeting assistant that records, transcribes, and summarizes calls — forever-free plan.
An AI assistant that answers questions and summarizes across your PDFs and documents — free tier with monthly pages.
An AI data analyst that lets you chat with your data to analyze and visualize it — free tier.
💻 Coding
An open source agent that helps you write code in your terminal, IDE, or desktop.
Open-source AI code assistant for VS Code and JetBrains — connects to any LLM (local or cloud) for autocomplete, chat, and inline edits.
AI pair programming in your terminal. Maps your codebase, edits multiple files at once, and commits changes with git — works with local or cloud LLMs.
A self-hosted, open-source AI coding assistant with a VS Code extension — runs models locally with no data leaving your machine.
An open-source autonomous coding agent for VS Code — plans tasks, edits files, and runs terminal commands with your approval, using any LLM you connect.
An open-source AI coding agent for VS Code — a whole dev team of AI agents that read, write, and run code in your editor using your own API keys.
An open-source, on-machine AI agent from Block — automates engineering tasks end to end and works with any LLM provider.
An open-source AI agent that brings Gemini to your terminal — code, debug, and automate with MCP server support and a generous free tier.
An open-source, high-performance code editor written in Rust with built-in AI assistant and real-time multiplayer editing.
An open-source AI coding agent for the terminal — access 300+ models via OpenRouter with a free daily token tier.
An AI-first code editor (VS Code fork) with deep codebase awareness, multi-file edits, and agent mode — free hobby tier plus paid Pro.
An agentic AI IDE with autonomous flows that read and edit across your codebase — free tier plus quota-based Pro.
AI code completion and chat across your editor and the CLI — free tier with monthly completions and chat included.
Free-forever AI autocomplete and chat across 70+ languages and most editors — unlimited completions on the free plan.
A privacy-focused AI completion assistant supporting 600+ languages with local model options — free tier plus Pro.
A modern, AI-powered terminal with agentic command generation and an intelligent command palette — free tier with monthly AI credits.
Google's asynchronous coding agent that tackles tasks in the background powered by Gemini — free tier with daily tasks.
An adaptive AI IDE with agent workflows and a token-based model — free tier plus a low-cost Lite plan.
An open VS Code AI agent extension with pay-as-you-go model access — free signup credits to start building.
🎙️ Audio
An open-source text-to-speech and speech-to-text model that delivers high-quality speech synthesis and recognition in a unified framework.
OpenAI's open-source speech recognition model that transcribes and translates audio in 99 languages with high accuracy — runs locally on CPU or GPU.
A lightweight, fast open-source TTS model that produces natural-sounding speech locally — small enough to run on CPU with near real-time performance.
A fast, local, open-source neural text-to-speech system — runs entirely offline on low-power devices down to a Raspberry Pi.
An open-source text-to-speech model built by inverting Whisper — fully open data and weights for natural, controllable voice synthesis.
An AI music generator that creates full songs with vocals from a text prompt — free tier with daily credits.
High-quality AI text-to-speech and voice cloning in many languages — free tier with monthly characters.
A fast, accurate speech-to-text API with streaming support — free tier with starter credits.
A speech-to-text API with real-time transcription and audio intelligence models — free tier with monthly hours.
🎬 Video
An open-source framework for agentic video generation, acting as a director, screenwriter, and producer.
Create videos programmatically using React.
Alibaba's open-source video generation model — produces high-quality short clips from text or image prompts and runs on consumer GPUs.
An open-source AI agent that automates creating short-form video content by generating scripts, sourcing clips, and editing Reels/Shorts/TikToks automatically.
Google DeepMind's text-to-video model producing cinematic clips — limited free access.
An AI video platform that creates avatar-presented videos in 120+ languages — free individual plan.
🖼️ Image
A highly powerful, modular, and node/graph-based GUI pipeline for Stable Diffusion, enabling custom AI image and video generation workflows.
The most widely used open-source web UI for Stable Diffusion — supports hundreds of models, ControlNet, inpainting, upscaling, and a massive extension ecosystem.
A professional open-source toolkit for AI image generation with a polished node-based canvas, workflow editor, and support for Stable Diffusion models.
An AI image generation suite with fine-tuned models and commercial-use rights — free tier with daily tokens.
An AI image generator known for accurate in-image text rendering — free tier with daily prompts.
An AI design tool that generates vector and SVG graphics as well as raster images — free tier with daily credits.
AI design assistants built into Canva for images, text, and layouts — robust free tier.
🧊 3D
Kimodo is a kinematic motion diffusion model trained on large-scale optical mocap data. It is controlled through text and constraints to generate high-quality 3D human and robot motions.
An open-source, AI-powered text-to-CAD application that runs entirely in the browser using WebAssembly to generate parametric 3D models from prompts and images.
✍️ Writing
An open-source, autonomous AI agent system designed for writing, auditing, and revising long-form novels.
An AI writing assistant for grammar, clarity, and tone with generative prompts — free tier with monthly AI prompts.
An open-source-backed grammar and style checker supporting 25+ languages — free tier.
An AI paraphrasing, summarizing, and grammar tool — free tier with per-use word limits.
A high-quality AI translator and writing assistant across many languages — basic free tier.
💰 Token
An optimization tool that enables running massive Large Language Models (like Llama 3 70B/405B) on consumer-grade hardware with extremely low VRAM.
An open-source tool for reducing LLM token usage and context window overhead — helps you fit more into fewer tokens without losing relevant information.

TL;DevTech // Edge & Local AI Hardware

HARDWARE STATUS HIGHLIGHTS
Nvidia RTX 3090 Available 24GB GDDR6X VRAM. A proven workhorse for local AI inference — runs 13B–70B quantized models at solid speeds. Great value on the used market for developers entering local AI.
Nvidia RTX 5070 Ti Available 16GB GDDR7 VRAM with next-gen Blackwell architecture. Exceptional FP8/FP4 throughput for local inference — rivals last-gen flagships at a lower price point.
Nvidia RTX 5090 Available 32GB GDDR7 VRAM. The flagship Blackwell GPU for local AI workloads — runs large quantized models with bandwidth and compute headroom to spare.
Nvidia RTX Spark Superchip Coming Soon Co-developed MediaTek ARM CPU + Blackwell GPU in a compact form. 128GB unified memory and 1 Petaflop FP4 performance lets developers natively run 120B parameter models on Windows with zero cloud latency.
Apple M6 Family Coming Soon Potential 2nm process with WMCM multi-chip packaging for memory bandwidth scaling. Designed to accelerate local open-source models using Apple's native MLX framework — no discrete GPU needed.