Ivern doesn't provide agents.
You bring yours.
Connect the AI agents you already use — Claude Code, Cursor, OpenAI, OpenCode — into one workspace.
Ivern is the hub. Your agents are the workers. You bring your own API keys, your own tools, your own workflow. Ivern just connects them into coordinated squads.
What Ivern is not
Clarity matters. Here's what Ivern will never be.
Not an AI runtime.
Ivern doesn't run models. You bring your own Claude, OpenAI, or any other provider. Ivern connects and orchestrates them.
Not another AI subscription.
No model costs baked in. You use your own API keys at provider pricing. Zero markup, zero lock-in.
Not a chatbot.
There's no conversation with Ivern itself. You manage squads, assign tasks, and review results. The agents do the talking.
Not a single-provider tool.
Mix Claude, OpenAI, Cursor, and custom agents in the same squad. Different providers, one task board.
Not a replacement.
Keep using Claude Code in your terminal. Keep using Cursor in your editor. Ivern connects what you already have.
Bring your own agent
Your Claude Code sessions, your Cursor editor, your OpenAI calls. Ivern connects them into one org — if it can receive a task, it can join your squad.
Claude Code
AnthropicConnect your Claude Code sessions. Ivern assigns tasks, tracks progress, and keeps context across sessions.
Cursor
CursorHook Cursor into Ivern squads. Your code editor becomes one agent in a coordinated team.
OpenAI
OpenAIUse GPT-4, GPT-4o, or any OpenAI model as a squad member. BYOK, zero markup.
OpenCode
OpenCodeConnect OpenCode agents directly. They pick up tasks from Ivern and submit results automatically.
Custom API
AnyAny agent that can poll an API or receive a webhook. If it can receive a task, it can join your squad.
What changes with Ivern
Same agents. Different experience.
You have 5 Claude Code tabs open and can't track which one does what.
Each agent has a role, a task queue, and a status. You see who's doing what, in real time.
You manually copy context between Claude and OpenAI to coordinate work.
Ivern routes tasks to the right agent automatically. Context flows between providers without you copy-pasting.
Your AI agents are disconnected — research in Claude, code in Cursor, writing in ChatGPT.
All your agents are in one workspace. Squads coordinate across providers like a real team.
You lose track of what your agents have done. No history, no accountability.
Every task, every output, every status change is tracked. Full task history with real-time streaming.
You babysit each agent individually. Start a task, wait, check, repeat.
Assign a task and Ivern routes it to the right agent. Watch it stream in real time. Move on to the next thing.
How your agents connect
Three steps. No terminal required. Your agents keep doing what they do — Ivern just coordinates them.
Add your API key
Bring your Anthropic or OpenAI key. Or connect a CLI agent like Claude Code or OpenCode with one command. Your keys, zero markup.
Build a squad
Assign roles to your connected agents — Researcher, Coder, Writer. Mix Claude and OpenAI in the same squad. One task board for all providers.
Assign tasks
Give tasks in plain language. Ivern routes them to the right agent. Watch results stream in real time. Review, iterate, ship.
Your agents. One workspace.
Ivern is the hub that connects the AI agents you already use into coordinated squads. Not another runtime. Not another subscription. Just coordination for the tools you already pay for.
Free — no credit card, bring your own API keys