MoriMori
Mori

Managed vs your own keys

Mori runs two ways: on API keys you bring, unlimited and free, or on managed frontier models we provision for subscribers. This page explains exactly how each works — including what happens when you hit a cap.

Two ways to power Mori

Every copy of Mori is the complete app — every tool, every surface, no feature gates. The only question is which models she thinks with, and there are two answers.

  • Your own keys — plug in an API key from any supported provider. Unlimited from our side; you pay your provider directly for exactly what you use.
  • Managed — Mori thinks with frontier models running on capacity we provision. No keys, no provider accounts, no per-token meter. One flat monthly price, and usage flows in sessions.

The free tier gives you both doors: unlimited use with your own keys, plus a fast-frontier managed allowance to try things without setting anything up. Subscriptions are how you make managed your main mode.

Your own keys — the free path

Bring a key from any of the nine built-in providers and everything works: computer use, browser, coding, terminal, voice, memory, MCP. Keys are stored locally and never leave your machine — Mori talks directly to your provider, nothing routes through us.

We put no limits on this path. Your provider bills you for the tokens you use, and that is the whole cost. If you are comfortable holding a key, the free tier is the full product, forever.

Managed — what a subscription adds

A subscription removes the key entirely. Mori runs on frontier models we provision and tune per tool — vision-strong models for computer use, fast ones for chat and quick steps. You never touch a provider dashboard or watch a per-token bill.

What a tier buys is capability and room to work: which model classes Mori reaches for, and how much she can do before a session refills. Higher tiers reach stronger model classes and carry more usage — nothing else about the app changes.

How usage flows: 5-hour sessions

Managed usage flows in 5-hour sessions that refill on their own. Your first message opens a session; five hours later it resets, whether you used a little or a lot. There is nothing to top up and nothing to meter — heavier tiers simply fit more work into each session.

  • Free and Plus also carry honest daily limits, sized so steady personal use rarely meets them.
  • Tiers scale relatively: Pro carries roughly 2x the usage of Plus, Studio roughly 2x Pro, and Max is the most room we offer.

What each tier reaches

  • Free — $0 — fast frontier models on a daily managed allowance, and unlimited use with your own keys.
  • Plus — $25/mo — fast frontier models with far more room, plus deeper-reasoning bursts when a step needs the extra depth.
  • Pro — $50/mo — advanced models as the everyday standard, on every task. ~2x the usage of Plus.
  • Studio — $100/mo — everything in Pro, plus our most powerful models in windows for the hardest work. ~2x the usage of Pro.
  • Max — $200/mo — everything first: new models on day one, top models through the day, our highest usage.
  • Custom — for teams that need more than Max; write to support@studiomori.ai.

Hitting a cap

Caps pause; they never lock. When a session or daily limit fills up in managed mode:

  • Managed usage pauses until the session resets — a few hours at most.
  • Or you switch to your own keys instantly, mid-conversation, and keep working with no interruption.
  • The app itself never stops. Every tool stays available the entire time.

There is no overage billing and no surprise invoice. A cap can only ever cost you a short wait or a key switch — never money.

Which should you pick?

If you already hold provider keys and like paying per token, the free tier is the complete product. If you want Mori to just work — no keys, no meters, frontier models tuned per task — pick the subscription whose model class matches your work: Plus for everyday speed, Pro for advanced depth as standard, Studio for top-model windows, Max for everything first.

The full tier-by-tier comparison lives at /pricing.