Evonic's blue jellyfish mascot

Agentic Infrastructure
for Organization

Transform your organization with agentic AI, making every workflow faster, smarter, and sharper at scale.

Your Models. Your Rules. Your Swarm.Scroll to explore

A HOME FOR EVERY AGENT

One place to
bring it together.

Create agents, choose their models, manage tools, and follow their activity from one workspace.

QUICK START Full guide

Install everything in one command.

$ curl -fsSL https://evonic.dev/install.sh | bash
Python 3.10+ Local or cloud LLM
THE EVONIC WORKSPACEEverything, connected.

A FRAMEWORK WITH ROOM TO GROW

Start with an agent.
Build a system around it.

Agent identity

A purpose. A personality. Possibilities.

BUILT FOR TRUSTED AUTONOMY

Memory that lasts.
Reach that goes further.

EVOMEM · PERSISTENT KNOWLEDGE

Remember more.
Retrieve what matters.

Give agents lasting memory from your Markdown files. Evomem combines keyword search, vector retrieval, and a knowledge graph to find relevant context without an LLM call at query time. Your knowledge stays easy to edit, back up, and version.

Persistent memoryHybrid searchMarkdown files
Explore Evomem

EVONET · ANYWHERE EXECUTION

Your agents.
Your infrastructure.

Put agents to work on laptops, servers, and edge devices. Evonet connects each device to Evonic through an outbound tunnel, so agents can run commands and work with files without SSH access, port forwarding, or a public IP.

Outbound connectionNo public IPAuto reconnect
Explore Evonet

BRING THE WORK TO YOUR AGENTS

Meet your workflow
where it is.

Talk in familiar channels

Web, Telegram, WhatsApp, and Discord.

Conversations connected to agent capabilities.

Keep recurring work moving

Use cron-based triggers, recurring tasks, and reminders.

Manage the system through the dashboard or CLI.

THE PEOPLE BEHIND EVONIC

Built in the open.
Better together.

Meet the contributors helping Evonic grow. Every fix, feature, and fresh perspective makes a difference.

YOUR NEXT CHAPTER STARTS HERE

Make room for
your first agent.

Choose a model endpoint. Set up Evonic. Start building.

Get started with Evonic Python 3.10+ · Local or cloud models