Plans, Limits & Billing
Which Claude plan do you need? Here's how to think about it — without quoting prices that would be stale next month (check the official pricing for those).
- Match a tier to a purpose instead of to a price
- See which capabilities are actually gated by plan — and which are not
- Know what hitting a usage limit feels like, and the cheapest ways to stop hitting it
- Stop confusing an app subscription with API billing
The tiers, by purpose
| Tier | Roughly for |
|---|---|
| Free | Trying Claude, light personal use |
| Pro | Regular individual use — more usage, more capability |
| Max | Heavy individual users who hit Pro limits |
| Team | Small teams — shared workspace and management |
| Enterprise | Orgs needing admin, security, scale, and controls |
(Exact names and structure can change — confirm on the pricing page.)
What's actually gated by plan
Most of the time the answer is usage, not features — which is why upgrading for one capability is usually the wrong reason. These are the exceptions worth knowing, as of October 2026:
| Capability | Free | Pro | Max | Team | Enterprise |
|---|---|---|---|---|---|
| Artifacts, code, charts | ✅ | ✅ | ✅ | ✅ | ✅ |
| File creation (docx/pptx/xlsx/pdf) | ✅ | ✅ | ✅ | ✅ | ✅ |
| Docs, Slides, Design | ✅ | ✅ | ✅ | ✅ | ✅ |
| Dashboards (beta) | ❌ | ✅ | ✅ | ✅ | ✅ |
| Motion (beta) | ❌ | ❌ | ❌ | ✅ | ✅ |
| Org-wide admin controls over the above | — | — | — | ✅ | ✅ |
Two patterns fall out of that table, and they generalize:
- Docs, Slides and Design launched in beta on paid plans and reached Free on October 8, 2026, once they left beta. Dashboards and Motion are at the earlier stage of the same journey. If a beta you want is above your tier, waiting is often the right call.
- A capability that is on by default on Pro, Max and Team is frequently off on Enterprise until an owner enables it in organization settings — and owners can narrow it to specific groups with custom roles. On Enterprise, "my plan includes it" and "I can use it" are different questions.
- Some capabilities need switching on even on individual plans — file creation lives behind Settings → Capabilities on Free, Pro and Max. Before concluding your plan lacks something, check the toggle.
What "limits" feel like
Consumer plans cap how much you can use in a window (messages and usage that reset over time), and higher tiers raise those caps. When you hit a limit you're typically asked to wait until it resets or upgrade — your work isn't lost, you just pause.
Heavier models and longer conversations consume your allowance faster. And since the September 2026 merge, one more thing is true: a casually-phrased request in the ordinary chat box can start a long agentic run that searches the web, executes code and writes files. The cost signal moved from the door to the prompt, so it pays to be explicit in both directions.
Keep a request cheap when you only want an answer
Answer from this conversation only. Do not search the web, do not run code, and do not create any files or artifacts. Two paragraphs.
Four habits that buy back more allowance than an upgrade does — all of them variations on context management:
- Every turn re-sends the whole conversation. A long thread you've drifted away from is pure overhead — see Tokens & Context.
- The prompt above is worth saving as a snippet. Most questions do not need a web search and a code run.
- Reaching for the most capable model for a quick rewrite burns allowance for nothing. See Models & Pricing for the current line-up.
- Three small corrections in one message cost roughly one turn's context; three separate messages re-send the thread three times. The exception is artifacts, where one change at a time lands more reliably.
How to choose
- Learn what you actually use Claude for before paying for a shape of usage you're guessing at.
- Not when you hit a limit once — when the interruption is costing you more than the subscription.
- This is a usage decision, not a capability one.
- Shared workspace, admin controls, compliance posture — and the two org-gated betas. The trigger is colleagues and governance, not raw volume.
Apps vs API billing — different things
- App plans = a flat subscription with usage limits (this page).
- API billing = pay per token for what you build. Different model, different page.
They do not share a balance, a limit, or a bill. Paying for Max does not give you API credit, and API spend does not raise your chat limits. Claude Code can draw on either, depending on how you authenticate it.
Quiz
Check yourself
0/3- Plans mostly gate usage, not features. Upgrading for a single capability is usually the wrong reason — check the toggle first.
- The current exceptions: Dashboards is beta on Pro and up, Motion is beta on Team and Enterprise only. Docs, Slides, Design and file creation are on every plan including Free.
- On Enterprise, "included in my plan" and "available to me" are different questions — owners enable capabilities per organization, and can narrow them to groups.
- Hitting a limit pauses you, it doesn't lose your work. Fresh chats, an explicit no-search instruction, and a right-sized model buy back more allowance than an upgrade.
- App subscriptions and API billing are separate systems with separate balances. Max gives you no API credit.
Next
- Getting Started with Claude.ai — if you're still deciding what you'd use it for
- Tokens, Context & Pricing (API) — the other billing model, explained
- Models & Pricing — the current line-up, so you can right-size the model
- Privacy & Data Handling — what differs between consumer and business tiers