Invite-only, while we're in beta.
Morchea is early, and we're onboarding deliberately — a small group at a time, so we can support each one well. Below are the plans we're opening with; every one is invite-only for now, with no charge for the engine yet. Request access and we'll be in touch.
Get on the list.
Tell us where to reach you. We review requests in small batches and send invites as we open up room.
No spam, no waitlist games. Just a note when there's room.
Pick the colony that fits how you work.
Two plans to start, with a third on the way. Every plan is invite-only while we're in beta — and the engine is free during beta. Most of your cost is just intelligence: models you already pay for, or credits you top up — your choice.
For fast-moving pros — designers, operators, marketers who run on AI but don't live in code.
Private beta · pricing soon. You bring the model spend.
- Your named Maestro & the full colony of seats
- Cockpit on the web — fully managed cloud
- Hands-free voice, the Council & the Harvest
- Pay with Morchea credits — no keys to wrangle
- Persistent memory & governed spend caps
For builders and vibe-coders — shipping real things, juggling projects, switching models all day.
Private beta · pricing soon. Bring your own keys.
- Everything in Go-getter
- Your own keys — cloud, or cloud + local models
- Route heavy work through your coding CLIs
- Many projects at once — drive one seat or the room
- Native macOS desktop app + the auren CLI
- Replay, fork & forecast every weave
For colonies that work together — A2A across your whole organization.
Not yet — but you can get in line.
- Everything in Founder
- Multiple colonies, one organization
- A2A — your Maestro speaks to your teammates'
- Shared Rooms, Harvest & memory
- Roles, admin & org-wide spend controls
You mostly pay for intelligence you already buy.
Morchea conducts the models you choose — or runs them for you on credits. Either way, most of your cost is just model usage, and you decide how much of it runs free.
Free to run
Point routine, high-volume work at open models on your own machine. Private by default, and nothing to meter.
$0 per run
Already paid for
Route heavy work through coding CLIs and subscriptions you already hold, so it runs under a plan you've bought.
No extra metering
Pay for what's used
For frontier-API work, you see the spend down to the task — and a hard cap means it can never surprise you.
Transparent, capped
Prefer not to manage keys at all? Top up with Morchea credits and we handle the model spend for you — capped and transparent. When we leave beta, paid plans will sit on top of this, priced for the orchestration and never hidden.
Before you ask.
Morchea works for hours or days at a time and acts on your behalf — so we'd rather onboard a few people properly than open the doors and disappoint a crowd. Small batches let us support each user and keep the system trustworthy.
The Morchea engine is free during the private beta. The only cost is your own model usage — and you control how much of that runs on free local models versus paid APIs.
Only if you want to. On the credits plan, we run the models for you — nothing to connect. If you'd rather bring your own (frontier APIs, your coding CLIs, or open models on your own machine), Morchea orchestrates those instead — and you're never locked to one provider. Either way, we set it up with you during onboarding.
Hand it a real job and the colony runs the whole thing — research and write a report, design and ship a website, build and debug an app, sweep your inbox and draft replies, or convene a multi-model Council on a hard call. You give the goal; it plans the work, does it, checks it, and brings back finished output — not just another chat.
You stay in control. Morchea works on its own for long stretches, but anything consequential — sending, publishing, or spending past a cap — pauses for your approval first. You set the guardrails, watch every step in a live log, and can step in anytime.
Your memory and your work are yours. Sensitive, high-volume work can run entirely on local models that never leave your machine, and you decide what each run is allowed to touch.
We're opening access in stages rather than against a fixed date. Request an invite and you'll be among the first to know as we widen the doors.