How much does Claude Code cost in 2026?

As of October 2026, the cheapest way to get Claude Code is the Claude Pro plan at $20 per month billed monthly, or $200 a year billed up front. Heavier users move to Max at $100 or $200 per month. Teams pay per seat: $25 per member per month on a Team Standard seat billed monthly ($20 billed annually), or $125 for a Premium seat ($100 annually). Enterprise and the pay-as-you-go API price usage at per-token API rates instead of a flat allowance. The free Claude plan does not include Claude Code.

The price tag is only half the answer. Every subscription has usage limits that reset every five hours and every week, shared with ordinary Claude chat. What you really buy is how much work fits inside them. Below you will find every plan, what the limits mean and how to pick without overbuying.

Prices change often, so check Anthropic's pricing page before you buy; prices here are USD, before tax. New to the tool? Start with what Claude Code is and what it does.

Which plans include Claude Code

Anthropic's setup documentation is blunt: "Claude Code requires a Pro, Max, Team, Enterprise, or Console account. The free claude.ai plan does not include Claude Code access." Here is every option as of October 2026.

Team plans run from 2 to 150 seats, and you can mix Standard and Premium seats on one plan. Confirm the Enterprise figure on the pricing page before you budget around it.

Pro and Max: plans for one person

Pro includes Claude Code in the terminal, the IDE extensions, the desktop app and the web, and shares one allowance with Claude chat. An hour of drafting in chat comes out of the same pool your Claude Code session uses.

Max comes in two sizes. Max 5x at $100 a month gives "five times the Pro plan's per-session usage allowance," and Max 20x at $200 gives 20 times. Max is "currently available as a monthly subscription only," so there is no annual discount. Fable, the highest-priced model in the API table below, is handled differently: Max can spend up to 50% of its weekly limits on Fable at no extra cost, while Pro pays for Fable with usage credits.

Team and Enterprise: plans for a business

Many older articles say Team Standard seats don't include Claude Code. That stopped being true on January 16, 2026, when Anthropic began "including Claude Code access with every Team plan standard seat." Anthropic's support article now says "Claude Code is included with every Team plan seat."

The difference between seats is usage. A Standard seat gets 1.25x Pro's per-session allowance and a Premium seat gets 6.25x. Give Premium seats to the people who use Claude Code all day. Each member's allowance is their own, resets on five-hour and weekly windows, and is shared with their Claude chat and Cowork use.

Team also adds SSO, server-managed settings the admin sets for everyone, the analytics dashboard, and Code Review (automatic pull request review, Team and Enterprise only). Code Review is a research preview billed separately through usage credits; Anthropic says each review averages $15-25. Team data falls under Anthropic's Commercial Terms: Anthropic does not train models on code or prompts sent to Claude Code unless the customer opts in. On Free, Pro and Max, model training is a setting you can turn off.

Enterprise works differently. New and self-serve Enterprise plans have a single seat type that includes Claude, Claude Code and Cowork, but the seat includes no usage. Everything is billed at standard API rates, and "there are no per-seat usage limits." Enterprise adds SCIM, the Compliance API and, for qualified accounts, zero data retention arranged separately with Anthropic.

Paying by the token: the API and cloud providers

With an Anthropic Console account, "Claude Code consumes API tokens at standard API pricing." A token is a small chunk of text; you pay for what Claude reads (input) and writes (output). October 2026 list prices:

Cached context is much cheaper (an Opus 5.5 cache hit is $0.20 per million tokens); full rates are on the API pricing page. Anthropic also notes that Claude 4.7 and later models use a tokenizer that "produces approximately 30% more tokens for the same text," so older estimates may run low.

Anthropic's cost documentation says that "across enterprise deployments, the average cost is around $13 per developer per active day and $150-250 per developer per month, with costs remaining below $30 per active day for 90% of users." That is a published average, not a quote for your team.

Claude Code also runs through Amazon Bedrock, Google Cloud's Agent Platform or Microsoft Foundry, "billed per token to your cloud account." Some features need a Claude subscription and don't work with Console keys or those providers, including cloud sessions, routines, Remote Control and Claude Code in Slack.

What usage limits mean in practice

"Both Pro and Max plans have a five-hour session limit and a weekly limit," and "these limits are shared across Claude and Claude Code." When you hit the five-hour limit, you wait for the window to reset or pay for more.

Anthropic does not publish limits as hours or prompts. It publishes multipliers: Max is 5x or 20x Pro, Team seats are 1.25x or 6.25x Pro. Anyone promising "X hours per week" is guessing. How fast you use your allowance depends on:

• The model. Bigger models cost more per token. Claude Code defaults to Opus 5.5 on Pro, Max, Team, Enterprise and the API, and you can switch with the /model command.
• The size of the job. A session that reads hundreds of files uses far more than a one-page fix.
• Parallel work. Several sessions at once, subagents or agent teams multiply usage. Anthropic says agent teams use approximately 7x more tokens than standard sessions when teammates run in plan mode.

Run /usage inside Claude Code to see plan usage limits and a breakdown of what counts against them. Anthropic notes that the dollar figure it shows for a session "is intended for API users."

Usage credits and the API credits that don't apply

If you hit your limits on Pro or Max, you can buy usage credits. They are "billed at standard API rates," apply to both Claude chat and Claude Code terminal usage, and you can set a monthly spend limit. In Claude Code, the /usage-credits command opens the right settings page, or sends a request to your admin on Team and Enterprise.

Since October 7, 2026, Max and Team plans also come with monthly API credits: $100 on Max 5x, $200 on Max 20x, and on Team $20 per Standard seat or $100 per Premium seat, pooled and capped at $500. These credits do not cover "interactive Claude Code in the terminal, IDE, desktop, or web." Team credits "can't be used in Claude Code or the Claude apps." The details are in Anthropic's article on monthly API credits.

One more trap: if an API key is set in your computer's environment, "Claude Code will use this API key for authentication instead of your Claude subscription," so you would be billed per token. Check which account you are using.

How to pick a plan

1. Count the people who will run sessions, not the people who will read results. A marketer who reviews a pull request doesn't need a seat.
2. Start one or two people on Pro or Team Standard for a short pilot on a real but low-risk job, working in git on a copy of the site.
3. Check /usage weekly. If limits interrupt work, upgrade that one person to Max or a Premium seat.
4. Choose Team once more than one person uses it for company work, so you get admin settings, SSO and commercial data terms.

Our 30/60/90-day rollout plan covers the pilot in detail.

Hypothetical monthly budgets for a small team

These are hypothetical examples with round numbers, using October 2026 list prices and monthly billing. They are not recommendations and they don't include usage credits you may choose to buy.

The subscription is rarely the biggest cost. People's time writing clear tasks, reviewing every change and testing before release usually is. Our guide to measuring the ROI of Claude Code shows how to count that honestly. If you plan to hire help, agency or consultant fees are separate from what you pay Anthropic; see what a Claude Code consultant costs.

Keeping costs under control

• Set spend limits. On Team and Enterprise, admins who turn on usage credits can set spend limits "at the organization, group, or individual member level."
• Match the model to the job. Smaller models cost less per token; try one on simple, repetitive edits and compare the results.
• One task per session. Clear the session between unrelated jobs so Claude isn't carrying old context. A short project file of instructions also saves Claude from rediscovering the same facts each time; our copy-ready CLAUDE.md templates are a starting point.
• Plan before editing. Plan mode lets Claude research and propose changes without editing files, so you catch a wrong approach before it burns through usage.
• Avoid paying twice. If your team also pays for an AI editor, compare what you get. Our guides to Claude Code vs Cursor and Claude Code vs GitHub Copilot explain how each bills for Claude models.

Conclusion

For most small businesses, Claude Code costs $20 a month for one person on Pro, rising to $100 or $200 for daily heavy use on Max, and $25 per member per month (or $20 billed annually) on a Team Standard seat. Start small, read your usage, upgrade only the people who hit limits, and budget for the review time every change needs before it ships. If you want help choosing plans, setting permissions and building a review process for your team, our Claude Code consulting team can set it up with you. More guides are in our Claude Code Guide.

Frequently asked questions