Free & Open Source AI Coding Agents

Zero-leak secret interception, deterministic Shannon entropy tokenization, AST shell command sandbox, and CI/CD security gate for Claude Code, Cursor, and autonomous AI agents.

Get up and running with Llama 3.3, Mistral, Qwen, and other large language models locally with a simple CLI and REST API.

An open-source LLM app development platform combining AI workflow, RAG pipeline, agent capabilities, and model management.

A high-throughput and memory-efficient inference and serving engine for LLMs with PagedAttention.

User-friendly AI interface designed for self-hosting with complete Ollama and OpenAI-compatible API support.

Command-line orchestrator for AI coding agents: runs Claude Code, Codex CLI, and Gemini CLI in parallel git worktrees with Kanban task state, typed merge flows, and per-agent skills.

Local, OpenAI-compatible REST API for running LLMs, audio transcription, image generation, and embeddings on commodity CPU/GPU.

The all-in-one Desktop & Docker AI application with full RAG pipeline, multi-agent workspaces, and complete privacy.

Drag & drop user interface for building customized LLM applications, autonomous agents, and RAG pipelines.

Dynamic visual framework for building, testing, and deploying multi-agent AI and RAG applications.

Open-source desktop alternative to ChatGPT that runs 100% offline on local hardware with zero data leakage.

Enhanced, self-hostable open-source AI chat platform unifying OpenAI, Anthropic, Gemini, and local LLMs.

Open-source prompt and automation framework for augmenting humans using AI patterns and modular CLI tooling.

AI pair programming in your terminal: edit code in local git repos with Claude 3.5 Sonnet, GPT-4o, or local LLMs with auto-commits.

Open-source platform for software development agents that can write code, fix bugs, run bash commands, and browse documentation autonomously.

Leading open-source AI code assistant inside VS Code and JetBrains: bring your own models, embeddings, and context providers.

High-performance LLM inference in pure C/C++ with minimal setup and state-of-the-art quantization across Apple Silicon and GPUs.

Call 100+ LLM APIs using the standardized OpenAI input/output format with load balancing, proxy, and spend tracking.

100% private, local document question and answering engine using open LLMs without internet connection or data leaks.

Open-source ecosystem of consumer hardware-optimized LLMs running locally with desktop client and Python bindings.

Self-hosted AI coding assistant alternative to GitHub Copilot with local model execution and editor plugins.

Open-source, personal AI second brain that searches, reasons over, and answers questions from your personal docs and notes.

Data framework for connecting enterprise custom data sources and knowledge bases to Large Language Models.

Standard open-source framework for building context-aware reasoning applications, agents, and multi-agent systems.

Structured data extraction from LLMs powered by Pydantic with validation, retries, and streaming support.

Platform for customizing AI models directly on top of real-time enterprise data engines using SQL queries.

The AI-native open-source embedding database with simple Python and JavaScript client SDKs.

Vector similarity search engine with extended filtering support, payload indexing, and high-performance Rust core.

Cloud-native open-source vector database built for massive scale similarity search and AI vector retrieval.

Open-source vector database that stores both objects and vectors, allowing for seamless hybrid search.

Transformer-based text-to-audio model created by Suno that can generate highly realistic speech, music, and ambient noise.

High-performance inference of OpenAI's Whisper automatic speech recognition in pure C/C++ without external dependencies.

The most powerful and modular stable diffusion GUI and backend with graph/nodes-based workflow interface.

Privacy-respecting open-source metasearch engine ideal for grounding AI models and retrieval-augmented generation.

Next-generation open-source RAG engine based on deep document understanding to extract structured data from complex files.

The memory layer for personalized AI assistants and autonomous agents, improving retention over time.

Open-source, local-first code interpreter that lets LLMs run Python, JavaScript, and Bash code on your computer.

Hugging Face's production-ready toolkit for deploying and serving Large Language Models at scale with Tensor Parallelism.

Fast serving engine for large language models and vision-language models with RadixAttention for multi-turn caching.

NVIDIA's specialized library providing state-of-the-art acceleration and optimized execution for LLM inference.

Fast inference library for running local LLMs on modern NVIDIA GPUs with EXL2 quantization.

Distribute and run LLMs with a single multi-GB binary that runs across Windows, macOS, Linux, and FreeBSD.

High-throughput, memory-efficient LLM inference engine supporting many architectures and quantizations.

Autonomous AI agent platform empowering developers to create, test, and deploy customized AI agent chains.

Cutting-edge framework for orchestrating role-playing, autonomous AI agents to collaborate seamlessly.

Microsoft's multi-agent conversation framework for building advanced AI agent workflows and collaborative task execution.

Multi-agent framework assigning distinct software company roles (Product Manager, Architect, Engineer) to LLMs.

Virtual software company simulated by communicative agents to complete full lifecycle software development.

Dev-first open source autonomous AI agent framework with GUI, agent telemetry, concurrent agents, and marketplace.

Finetune Llama 3.3, Mistral, and Qwen 2x-5x faster with 70% less memory using optimized hand-written GPU kernels.

Post-training framework for fine-tuning hundreds of open-source language models with standard YAML configurations.

Unified web UI and CLI for easy fine-tuning and evaluation of 100+ LLMs and multi-modal models.

Activation-aware Weight Quantization package for 4-bit LLM quantization with 3x speedup.

Easy-to-use LLMs quantization package with user-friendly APIs, based on GPTQ algorithm.

Accessible 8-bit and 4-bit optimizers and quantization functions for deep learning on CUDA and ROCm.

Hugging Face's state-of-the-art Parameter-Efficient Fine-Tuning library (LoRA, QLoRA, Prefix Tuning).

Transformer Reinforcement Learning library with Supervised Fine-Tuning (SFT), DPO, and PPO training loops.

Evaluation framework for your Retrieval Augmented Generation (RAG) pipelines with faithfulness metrics.

The open-source LLM evaluation framework like Pytest for LLM applications with 14+ benchmark metrics.

NVIDIA's toolkit for easily adding programmable guardrails and safety bounds to LLM-based conversational apps.

A guidance language for controlling large language models with interleaved generation, prompting, and constrain.

Structured text generation library with regular expressions, JSON schemas, and context-free grammars.

Test your prompts, agents, and RAGs against red-teaming attacks, prompt injections, and regression suites.

The ubiquitous browser interface for Stable Diffusion based on Gradio library with extensive extensions.

Professional generative AI engine and canvas for creative visual artists and digital designers.

Focus on prompting and generating without complex parameter tuning, inspired by Midjourney simplicity.

Visual instruction tuning toward large multimodal models: chat and reasoning over high-res images.

Meta's foundation model for image segmentation, allowing promptable cutouts of any object with one click.

Ultralytics YOLOv8 for real-time object detection, image segmentation, pose estimation, and classification.

Awesome multilingual OCR toolkits supporting 80+ languages, table recognition, and document structure analysis.

Lightweight, open-weight text-to-speech model producing studio-grade human voices with only 82M parameters.

Fast, local neural text to speech system that sounds great and is optimized for Raspberry Pi and low-power hardware.

A generative text-to-speech model specifically designed for conversational scenarios and dialogue synthesis.

Reimplementation of OpenAI's Whisper model using CTranslate2, delivering up to 4x faster transcription.

Open-source vector similarity search extension for PostgreSQL, supporting exact and approximate nearest neighbor.

Meta's library for efficient similarity search and clustering of dense vectors, indexing billions of vectors on GPUs.

Developer-friendly, serverless open-source vector database for multi-modal AI with zero-copy storage.

The open big data serving engine: store, search, organize, and machine-learn over large vector and text datasets in real-time.

Spotify's C++ library with Python bindings to search for points in space that are close to a given query point.

Pioneering AI-powered task management and autonomous agent execution system that builds on task completion loops.

Open-source code generation system integrating dynamic execution feedback and compiler outputs for iterative code refinement.

A query language for programming Large Language Models with typed output constraints and distribution control.

Intel's advanced weight-only quantization algorithm for low-bit LLM optimization without accuracy degradation.

Autonomous software engineering agent that turns LM models into autonomous agents solving GitHub issues.

Stanford framework for programming—rather than prompting—foundation models with automatic prompt optimization.

An open platform for training, serving, and evaluating LLM-based chatbots, the core engine behind LMSYS Chatbot Arena.

Semantic cache for LLM queries to cut costs and speed up response time up to 10x using vector similarity.

Open-source MCP server connecting AI assistants to Dubai and UAE public APIs — real estate, transport, statistics, education, health, and business data via Dubai Pulse, RTA, DLD, KHDA, and DHA. Runs locally over stdio.

Free, MIT-licensed plugins for Claude Code covering maintained codebase maps, project rules, coding-work orchestration, capability discovery, model routing, and local observability. Developers can install individual plugins rather than the entire collection; Claude or model-provider usage may have its own costs.

Open-source project workspace for teams and AI agents, with task delegation, persistent sandbox workspaces, and deliverable review. Self-hosting uses separately configured model-provider credentials.

Kubernetes CLI to manage your clusters in style, observe real-time events, and troubleshoot workloads.

Run your GitHub Actions locally inside Docker containers for fast, offline CI/CD pipeline iteration.

Tool for exploring a Docker image, layer contents, and discovering ways to shrink your image size.

A simple terminal UI for both Docker and docker-compose, written in Go with the gocui library.

Vulnerability scanner written in Go that checks your dependencies against the Open Source Vulnerability database.

Multi pod and container log tailing for Kubernetes with regex filtering and color-coded stream output.

Faster way to switch between clusters and namespaces in kubectl with fuzzy search integration.

The package manager for Kubernetes: define, install, and upgrade complex Kubernetes applications.

Customization of Kubernetes YAML configurations without templates, using patch layers and overlays.

Run Kubernetes locally on Windows, macOS, and Linux with automated cluster provisioning.

Kubernetes IN Docker: tool for running local Kubernetes clusters using Docker container nodes.

Top-like interface for container metrics, real-time memory and CPU consumption for Docker and Kubernetes.

Ultra-low latency open-source LLM inference API (Llama 3.3, DeepSeek, Qwen) with a generous free developer tier.

Instant serverless API access to 100,000+ open-source AI models across NLP, vision, and speech with free monthly quota.

Search indexed YouTube transcripts by topic, with source links and timestamps. Retrieve captions and resolve people, channels and videos. Coverage is partial. A permanent Free tier is available; Premium transcripts and sponsor research use paid plans.

Deterministic social caption and hashtag generator API returning platform-ready JSON. Features a permanent keyless playground endpoint for developers and AI agents.

Developer-first email platform for sending transactional emails using clean React components.

Email marketing and transactional messaging platform with reliable delivery and webhook tracking.

Cloud-based email infrastructure platform with robust delivery APIs, template engines, and analytics.

Developer-first email delivery platform offering inbound routing, SPF/DKIM verification, and logs.

Full-featured, hackable Next.js AI chatbot built by Vercel using the AI SDK, Shadcn UI, and Neon Postgres.

Starter template for building full-stack LangChain applications with Next.js App Router and streaming responses.

Turn Python data scripts into shareable web applications in minutes with built-in charting and interactive widgets.