AI Agents,
Orchestrated.

Meet ExtendedLM. A comprehensive AI agent platform with multi-agent orchestration, workflow automation, and Model Context Protocol integration. Build intelligent systems that work.

Multi-Agent System

Orchestrate multiple specialized agents including Translation, Weather, RAG, PDF Reading, and Computer Use agents. Each agent excels at specific tasks.

Workflow Automation

Build complex workflows with visual editors and ReactFlow integration. Chain agents together for powerful automation pipelines.

MCP Integration

Seamless Model Context Protocol integration for filesystem access, playwright automation, and extensible tool support.

Specialized Intelligence

Powerful Agent Ecosystem.

ExtendedLM provides a rich ecosystem of specialized agents. Each agent is optimized for specific tasks, from translation to computer automation.


Available Agents

High-accuracy Translation, Real-time Weather Data, Document Search with RAG, PDF Analysis, Slide Generation, Computer Automation, and extensible MCP agents. Combine them to automate your workflows.

ExtendedLM Agent Ecosystem

Multi-Agent Ecosystem Overview

Visual Workflow Orchestration

Visual Workflow Orchestration

Visual Workflows

Orchestration, Simplified.

Build complex workflows visually with ReactFlow integration. Drag and drop agents, connect them with logic flows, and watch your automation come to life.


Dynamic Execution

ExtendedLM executes workflows intelligently. Chain multiple agents, handle errors gracefully, and maintain state across complex multi-step processes. All with a visual editor that makes complexity manageable.

Built for Real-World Use.

Flexible, extensible, and ready for production.

Any LLM Provider

Switch between OpenAI, Anthropic, Google, xAI, and local models without changing your code. Avoid vendor lock-in.

Visual Workflow Builder

Drag and drop agents to create complex automation. No coding required for basic workflows.

Fully Extensible

Add custom agents, tools, and integrations. MCP support lets you connect to any external system.

Developer First

Built for Developers.

ExtendedLM is designed with developers in mind. TypeScript strict mode, comprehensive testing with Vitest, and modern development practices.

  • ✓
    CLI Interface: Powerful command-line tools for workflow execution and testing.
  • ✓
    Web UI: Beautiful Next.js interface with real-time updates via Socket.io.
  • ✓
    Extensible: Add custom agents, tools, and workflows with clean TypeScript APIs.
Workflow Console
RUNNING
Translation Agent → RAG Agent
Processing 142 documents in workflow...
<Workflow>
  <Agent type="rag" />
  <Agent type="pdf" />
  <!-- Chain agents -->
</Workflow>

Powered by Integrated Components.

ExtendedLM integrates Gateway and Mate to deliver a complete AI platform.

Gateway

Unified LLM proxy that connects OpenAI, Anthropic, Google, xAI, and local models through a single API. Switch providers without code changes. Built in Rust with GPU acceleration and llama.cpp integration.

OpenAI Anthropic Google xAI llama.cpp
Learn More →

Mate

AI-powered browser automation with natural language commands. Real browser control via Docker + Chromium with live VNC monitoring. Execute complex web tasks with simple instructions.

Docker Chromium VNC Playwright
Learn More →

Complete AI Automation Platform

Flexible LLM Selection Choose the optimal model for each task through Gateway. Local models for privacy, cloud for power.

Intelligent Automation Orchestrate multi-agent workflows that combine browser automation with advanced reasoning.

Production Ready Enterprise-grade architecture with Docker isolation, comprehensive logging, and real-time monitoring.

Enterprise Ready.

Production-grade features for real-world applications.

Agent Ecosystem 10+ Agents
Translation, RAG, Computer Use...
Model Support 5+ Providers
Anthropic, OpenAI, Google, Ollama
Testing Coverage Vitest
Comprehensive Test Suite

Start building with ExtendedLM.

Open source and ready to deploy. Start building intelligent agent systems today.