Guide · updated 2026-10-14 · 9 min read

Claude Code Pricing (October 2026): Which Plan to Pick

Claude Code is included in every paid Claude plan: Pro at $20 a month ($17 a month billed annually), Max at $100 or $200 a month, Team seats from $20 to $125, and Enterprise at $20 a seat plus usage at API rates. You can also skip the subscription and pay per token through a Claude Console account, which is the right call for automation and the wrong one for most people who code with an agent all day.

I checked every number below against Anthropic's official pricing page and help center on October 14, 2026. Anthropic changes plans often, so treat that page as the source of truth and this guide as the explanation of what the numbers mean in practice.

The plans at a glance

All of this comes from the pricing page. Team is for groups of 2 to 150 people, and you can mix seat types. Max is monthly billing only.

What "usage" actually means for agent work

This is the part most pricing roundups gloss over. Subscriptions don't give you a token count. Per Anthropic's pricing FAQ, every plan has limits that reset on a rolling five-hour window, paid plans add weekly limits on top, and there is no fixed message count because it depends on model, context length and features.

Three things matter for daily agent use:

1. One pool for everything. Claude chat on web, desktop and mobile shares the same allowance as Claude Code. If you spend the morning in long chat threads, your afternoon coding session starts with less.
2. The IDE counts too. The Pro and Max help article confirms the VS Code, Cursor and JetBrains extensions draw from the same limits as the terminal.
3. Long sessions are expensive even when you type little. The costs docs explain that Claude Code resends the full conversation with every request, so a one-line question in a session open all day still draws usage for the whole history. Subagents, agent teams and scheduled /loop tasks all send their own requests on top.

When you hit a limit, you can wait, upgrade, or turn on usage credits, which keep you working at standard API rates. Usage credits are opt-in and you can set a monthly spend cap.

Plan by plan, for people who code with agents daily

Pro: fine for trying it, tight for daily driving

Anthropic's own Claude Code page describes Pro as suited to "short coding sprints in small codebases." That matches what I'd expect: enough to learn the tool and do focused tasks, not enough to leave an agent running on a big repo for hours. Pro also gets Opus, but the top-tier Fable models are only available on Pro through usage credits, per the plan comparison table.

Max 5x: my default recommendation

At $100 a month you get five times Pro's per-session allowance, priority access at busy times, and access to Fable within its own limit (Max plans get 50% of weekly limits for Fable, and the help center notes a separate weekly Fable cap). If you use Claude Code most days, this is where I'd start. Upgrade to 20x only after you've hit the weekly limit more than once in a normal week, not because you're anxious about it.

Max 20x: for parallel sessions and long-running work

$200 a month. Worth it if you routinely run several Claude Code sessions at once, use agent teams, or let tasks run for hours. The docs say agent teams use roughly 7x more tokens than standard sessions when teammates run in plan mode, so parallel workflows burn allowance fast.

Team: Standard for light users, Premium for your builders

Every Team seat includes Claude Code, according to the Team and Enterprise help article. Standard seats give "more usage than Pro," Premium seats give 5x a Standard seat. You also get SSO, central billing, admin controls and no model training on your content by default.

Anthropic's wording suggests a Premium seat lands in the same neighborhood as Max 5x, but exact allowances aren't published, so I wouldn't promise anyone parity. Practical pattern: put people who mostly chat on Standard and anyone using Claude Code daily on Premium. Owners can enable usage credits with org and per-user spend limits for the occasional heavy week.

Enterprise: a seat fee plus a meter

New Enterprise plans use a single seat at $20 per user per month, billed annually, and that seat includes no usage. Every token is billed at standard API rates. On self-serve Enterprise you buy credits up front, and when the shared balance hits zero, usage stops for the whole organization until an owner buys more. Set spend limits per person, or one person's long refactor drains everyone's budget.

Console (API): pay per token

You can sign into Claude Code with a Console account and pay standard API pricing. Current list prices per million tokens:

For a reality check on what that adds up to, Anthropic's costs documentation reports enterprise averages of about $13 per developer per active day and $150 to $250 per developer per month, with 90% of users staying under $30 per active day. That's the same range as Max 5x and Max 20x, except a subscription has a ceiling and the API doesn't.

Where the API wins: CI jobs, headless claude -p scripts, shared automation, and teams that need per-key spend reporting and workspace limits. Where it loses: an individual who codes all day and wants a predictable bill.

Two billing traps worth five minutes

Trap 1: the Max and Team API credits don't cover Claude Code

Max 5x now includes $100 a month in API credits, Max 20x $200, and Team up to $500 pooled ($20 per Standard seat, $100 per Premium seat). The API credits docs are explicit that these cover the Claude API, Managed Agents, the Agent SDK and the playground, but not Claude Code and not extra usage in the apps. They expire monthly with no rollover. Use them for side projects built on the API, not to extend your coding sessions.

Trap 2: a stray API key silently switches you to API billing

From the Pro and Max help article: if ANTHROPIC_API_KEY is set in your environment, Claude Code uses it instead of your subscription. To check and fix:

1. Run echo $ANTHROPIC_API_KEY in the shell where you start Claude Code. Also grep your .zshrc, .bashrc and any .env files your shell loads.
2. Remove or unset it for that shell.
3. In Claude Code, run /logout, then /login, and pick "Claude account with subscription."
4. Run /usage. Subscribers see plan usage bars; the dollar "Total cost" figure in the Session block is an estimate at list price and isn't relevant for subscription billing.

How to pick, in four steps

1. Start on Pro for a week if you're new to Claude Code. Watch /usage daily.
2. Move to Max 5x the first time you hit the five-hour limit mid-task more than a couple of times a week.
3. Move to Max 20x only if you hit the weekly limit in a normal week, or you run parallel sessions as a habit.
4. Use the API alongside, not instead, for CI and scripted runs. Give that work its own Console workspace with a spend limit.

For a team of five or more who need SSO and admin controls, Team with a mix of Standard and Premium seats is the default. Go to Enterprise when you need SCIM, audit logs or custom retention and you're comfortable budgeting for metered usage.

Make any plan go further

The same habits cut cost on the API and stretch limits on a subscription. These come from Anthropic's costs docs:

• /clear between unrelated tasks. It costs nothing; /compact itself is a large request.
• Default to Sonnet, switch to Opus with /model when the task needs it, and set model: haiku on simple subagents.
• Lower effort with /effort for routine work. Thinking tokens bill as output.
• Keep CLAUDE.md under about 200 lines and move workflow-specific instructions into skills. See how to write a CLAUDE.md file.
• Mind breaks. The prompt cache lasts an hour on a subscription but drops to five minutes once you're on usage credits, so the first message after lunch can reprocess your whole context.
• Check the /usage breakdown. It attributes recent usage to skills, subagents, plugins and MCP servers, and flags anything above 10%.

More of these in Claude Code best practices.

One trade-off to be honest about: limits on subscriptions are deliberately vague, and Anthropic reserves the right to add caps at its discretion. If you need a hard, predictable number for budgeting a team, API or Enterprise billing gives you that, at the cost of an open-ended meter.

Bottom line

As of October 14, 2026: Free has no Claude Code. Pro ($20) is for trying it. Max 5x ($100) is the sensible default for daily agent use, and Max 20x ($200) is for parallel or long-running work. Team Premium seats suit your heavy users; Enterprise and the API bill every token, which is great for control and bad for surprise invoices. Check the official pricing page before you buy, because it will change.

If this was useful, Agent Shipped covers setups like this in a free 4-minute read, twice a week, and you're welcome to subscribe. ```

Files are in this folder: - claude-code-pricing.md

FAQ

Is Claude Code included in the free Claude plan?

No. Anthropic's pricing page lists Claude Code as included in Pro, Max 5x and Max 20x but not Free. Team and Enterprise seats also include it.

Do the monthly API credits on Max and Team work in Claude Code?

No. The $100 (Max 5x), $200 (Max 20x) and up to $500 pooled (Team) monthly credits only cover the Claude API, Managed Agents, the Agent SDK and the playground. Claude Code and extra usage in the Claude apps are explicitly excluded.

What happens when I hit my Claude Code limit on Pro or Max?

You can wait for the five-hour or weekly window to reset, upgrade, or turn on usage credits, which bill at standard API rates. Usage is shared with Claude chat, so heavy chatting eats into your coding budget.

Is the API cheaper than Max for Claude Code?

For most daily users, probably not. Anthropic's own docs put average enterprise API spend at around $150 to $250 per developer per month, which is the same range as Max 5x and Max 20x, and the API has no ceiling unless you set one.

Why am I being charged API fees when I have a Pro or Max plan?

The usual cause is an ANTHROPIC_API_KEY environment variable. If it's set, Claude Code uses the key instead of your subscription and bills you at API rates. Unset it, then run /login and pick your Claude account.