A technical index of verified AI software, local model runners, autonomous agents, and automation frameworks.
Multi-agent framework enabling autonomous AI teams to role-play, collaborate, and execute complex goals
Orchestrates collaborative autonomous agents where each agent operates with defined roles, tools, and delegation rules. Seamlessly executes multi-step research, content pipelines, and operational tasks.
Autonomous open-weights agent framework with native function calling and tool execution
Built by Nous Research, Hermes Agent powers autonomous workflows using Hermes 3 and Llama 3 models. Features structured JSON output, native tool execution, persistent agent memory, and compatibility with local runners like Ollama and vLLM.
Build resilient, multi-actor LLM applications with cyclic computation and state graphs
Engineered by LangChain, LangGraph provides fine-grained control over agentic loops, human-in-the-loop approvals, and multi-turn state persistence. It is the enterprise standard for complex multi-agent architectures.
High-performance proxy to call 100+ LLMs in OpenAI standard format with spend guardrails
Lightweight gateway that standardizes API requests across AWS Bedrock, Azure, Anthropic, OpenAI, Vertex, and local vLLM instances. Features per-key budget limits, fallback routing, and team spend analytics.
Ultra-realistic AI voice cloning, conversational agents, and multilingual speech synthesis
Generates natural human voices across 29+ languages from plain text, featuring instant voice cloning, emotional nuance controls, and conversational low-latency voice agent APIs.
Open-source visual canvas for building customized LLM chains, autonomous agents, and RAG pipelines
Drag-and-drop orchestration UI for LangChain and LlamaIndex. Lets teams visually connect vector stores, memory, custom tools, and autonomous agent loops without writing boilerplate backend code.
Visual drag-and-drop workflow automation with multi-branch logic and AI connectors
Connect 1,500+ business applications without code. Automate customer data flows, CRM entries, and content publishing with built-in AI routing, error handling, and JSON parsing.
Fair-code workflow automation platform with native AI agent orchestration and 400+ nodes
Visual workflow automation platform for technical teams. Combines 400+ API connectors with native LangChain AI agents, vector database stores, JavaScript/Python code execution, and complete self-hosting freedom.
Autonomous coding agent in VS Code that executes terminal commands, edits files, and uses MCP
An open-source autonomous coding extension for VS Code. Pairs frontier models such as Claude 3.7 Sonnet, DeepSeek-V3, and GPT-4o with bash execution, file modifications, browser automation, and Model Context Protocol servers under strict human supervision.
The AI code editor built for codebase indexing, multi-file refactoring, and agentic editing
A specialized VS Code fork engineered for autonomous software engineering. Features Composer for multi-file code editing, deep semantic codebase indexing, terminal error fixing, and model access to Claude 3.7 Sonnet and GPT-4o.
Run frontier open-source LLMs locally on macOS, Linux, and Windows with GPU acceleration
Lightweight tool to bundle model weights, configurations, and quantization into portable Modelfiles. Exposes an OpenAI-compatible REST API and terminal CLI supporting Llama 3, DeepSeek-R1, Mistral, and Qwen 2.5 on Apple Silicon and NVIDIA GPUs.
Unified API gateway and intelligent model router for 200+ frontier and open-weights LLMs
Single OpenAI-compatible endpoint with automatic fallback routing, token-level pricing transparency, rate-limit pooling, and support for DeepSeek-R1, Claude 3.7 Sonnet, Llama 3.3, and Gemini 2.5.
High-throughput and memory-efficient LLM serving engine powered by PagedAttention
Developed at UC Berkeley, vLLM delivers up to 24x higher serving throughput than standard Hugging Face pipelines by managing KV cache memory with PagedAttention. Powers production model inference with continuous batching and speculative decoding.
Generate styled presentations, technical documents, and web pages from text prompts
Turns markdown outlines or text briefs into interactive slide decks, document hubs, and web pages with one-click theme styling and responsive layout formatting.
Conversational answer engine that searches the live web and delivers cited syntheses
Replaces traditional web search with instant, cited answers synthesized across live web pages, research papers, and technical documentations. Features Pro Search multi-step research modes.
Frontier reasoning, advanced coding, and interactive Artifacts from Anthropic
Anthropic's flagship family including Claude 3.7 Sonnet with hybrid reasoning and Claude 3.5 Sonnet. Excels at complex system architecture, nuanced technical writing, tool use, and interactive web prototyping.