Choosing a plan.
5 MIN READ · UPDATED 2026-08A short guide to picking the right tier without paying for capacity you don't use — and knowing when Max is genuinely the right move.
Most AI subscriptions push you toward the more expensive tier by default. LADLE's honest position is the opposite: most people are fine on Base ($20/mo), and if you're not sure whether you need Max, you don't need Max.
The default: Base
Base includes the same Claude Sonnet model, the same 200K context window, the same file handling, the same Projects, and everything else in the product. The only thing that changes across tiers is the usage cap.
For a typical daily user — sending 30-100 messages a day, attaching a file or two, using Projects for ongoing work streams — Base is not just enough, it's overprovisioned. You'll rarely see the cap.
When Max is defensible
Max 5x ($100/mo) is the right move when both of these are true:
1. You're regularly hitting the Base cap and it's blocking work (not just annoying). 2. Compressing your usage isn't practical — you can't split chats, prune Projects, or reduce file sizes without losing the workflow.
If only the first is true, look at Projects mastery and see if your usage pattern can be tighter. Most people who think they need Max are actually running one enormous Project that could be three focused ones.
Max 20x ($200/mo) is for people who genuinely spend 4+ hours a day in AI-assisted coding or research — the "second monitor is Claude" pattern. If that's you, Max 20x is not extravagant; it's honest pricing for how much inference you consume.
The signal to watch
Settings → Usage shows your rolling 30-day token consumption. If your line is trending toward the cap consistently — not one bad day, a sustained pattern for 2-3 weeks — the upgrade math changes.
If your line stays well under the cap even on your busiest days, no amount of "just in case" reasoning should push you off Base. You'd be paying for capacity you won't use, and the meal donation math still favors Base for people who don't actually need Max ($8/mo funding ~10 meals is a stronger per-dollar contribution than the marginal donation on Max tiers).
Switching is free
There's no lock-in. Upgrade takes effect instantly, prorated. Downgrade takes effect at the next cycle. If you try Max for a month and don't need it, you're not out anything except the difference in monthly cost — and you got the meals contribution from that month.
Team plans
Team plans are per-seat at the same tier structure. Every seat in a workspace is on the same plan tier — you can't split. If your team has genuinely different usage patterns (some heavy coding, some light chat), the cleanest structure is often multiple workspaces, or personal plans plus one team workspace for shared work. If you're not sure how to structure this, email us — small volume of these questions and we prefer to talk them through.
The one thing not to do
Don't upgrade "just to be safe" on your first day. You have no data yet. Use Base for a full billing cycle, watch your dashboard, and let the actual pattern make the decision for you.
- Start on Base. You have no usage data yet — let the actual pattern decide.
- Max is the right call only when you're consistently hitting the cap AND compression isn't practical.
- Switching is instant and prorated in both directions — no lock-in.
- Team plans are per-seat, workspace-level; different tiers require different workspaces.
- Don't upgrade 'just to be safe' — you'd be paying for unused capacity.