Usage, limits and cost

How Much Does Claude Code Cost Per Month? Pro vs Max vs API (2026)

Claude Code cost per month on Pro, Max 5x, Max 20x, Team and the API, with per-token prices, what drives your bill and how to check your spend.

On this page
  1. Claude Code pricing at a glance
  2. What each subscription plan gets you
  3. API pricing for Claude Code's models
  4. What Claude Code actually costs per month
  5. Subscription or API: which is cheaper?
  6. Ways you can pay more than you expect
  7. How to check what you're spending
  8. Prices in other countries
  9. FAQ

Claude Code costs $20 a month on the Pro plan ($17 a month if you pay $200 for a year), $100 a month on Max 5x and $200 a month on Max 20x, because it's included in every paid Claude plan and shares that plan's usage limits. If you sign in with an API key instead, there's no monthly fee: you pay per token, from $1 per million input tokens on Haiku 4.5 to $10 on Fable 5.1, and Anthropic says the average enterprise developer spends about $13 per active day, or $150 to $250 per month.

Every price and limit below comes from claude.com/pricing, the Claude API pricing page, the Claude Code cost docs and Anthropic's support articles, all checked on October 3, 2026. Prices change, so treat those pages as the final word.

Claude Code pricing at a glance

PlanPrice (before tax)Claude CodeHow much usage
Free$0Not includedNo Claude Code
Pro$20/month, or $17/month billed $200 yearlyYesAt least 5x Free per 5-hour session, plus a weekly limit
Max 5x$100/month (monthly only)Yes5x Pro per 5-hour session, plus a weekly limit
Max 20x$200/month (monthly only)Yes20x Pro per 5-hour session, plus a weekly limit
Team, Standard seat$25/seat/month, or $20 billed yearlyYesMore than Pro
Team, Premium seat$125/seat/month, or $100 billed yearlyYes5x a Standard seat
Enterprise$20/seat/month billed yearly, plus usage at API ratesYesMetered, with admin spend limits
API key (Claude Console)No fee, pay per tokenYesNo plan limit, only API rate limits

The Pro and Max prices come from claude.com/pricing and the support article What is the Max plan?, which lists Max 5x at $100 and Max 20x at $200 per month for web subscriptions. Mobile app prices can differ. Team is for 2 to 150 people.

Notice what the table doesn't say: how many tokens or messages each plan includes. Anthropic doesn't publish that. Its pricing FAQ says usage depends on "the length and complexity of your conversations, the model you choose, and the features you use, so there's no fixed message count." Our guide to the Claude Code 5-hour limit covers what that means day to day.

What each subscription plan gets you

Every paid plan includes Claude Code in the terminal and in supported IDEs (VS Code and its forks, and JetBrains IDEs), according to Anthropic's Pro and Max guide. The important detail is that your usage is one pool: chats on claude.ai, the desktop app and Claude Code all draw from the same limits.

Pro: $20 a month

Pro is the cheapest way to use Claude Code. You get Opus, Sonnet and Haiku, a 5-hour session limit and a weekly limit. The pricing page lists Fable on Pro as available through usage credits only, which means pay-as-you-go on top of the subscription.

Pro is enough if you use Claude Code for focused stretches and rarely see the "You've hit your session limit" message. If you see it most days, look at Max.

Max 5x and Max 20x: $100 and $200 a month

Max multiplies Pro's per-session allowance by 5 or 20. Both tiers are monthly only. Max also adds higher output limits, priority access at busy times, and up to 50% of the weekly limit on Fable, according to the plan comparison on claude.com/pricing.

The jump from Pro to Max 5x is five times the price for five times the session allowance, so the per-unit price doesn't drop. You're paying to hit the wall less often, not for cheaper tokens.

Team and Enterprise

On Team, each member's Claude Code usage draws from a per-seat allowance that resets on a rolling five-hour window and a weekly window, as the costs docs explain. Premium seats get five times a Standard seat. Enterprise is different: a $20 seat fee plus usage billed at API rates, so its cost behaves like the API rows below.

API pricing for Claude Code's models

If you sign in with a Claude Console account or set ANTHROPIC_API_KEY, Claude Code bills every request per token. These are the list prices per million tokens (MTok) from the API pricing page:

ModelInputOutputCache write (5 min)Cache write (1 hour)Cache read
Fable 5.1$10$50$12.50$20$0.25
Opus 5.5$4$20$5$8$0.20
Sonnet 5.5$2$10$2.50$4$0.20
Haiku 4.5$1$5$1.25$2$0.10

A few multipliers sit on top. Fast mode on Opus 5.5 costs $8 input and $40 output. US-only inference costs 1.1x. There's no premium for long context: a 900K-token request costs the same per token as a 9K one.

Which model you pay for matters. On the Anthropic API, Pro, Max, Team and Enterprise, Claude Code's default model is Opus 5.5, according to the model configuration docs. Switching routine work to Sonnet 5.5 halves the input and output price.

What Claude Code actually costs per month

On a subscription, your monthly cost is the plan price, plus any usage credits you choose to buy. On an API key it depends entirely on how you work. Anthropic's own numbers, from the costs docs:

  • Across enterprise deployments, the average is about $13 per developer per active day and $150 to $250 per developer per month.
  • For 90% of users, costs stay below $30 per active day.
  • Background work such as summarizing past conversations for claude --resume typically uses under $0.04 per session.

Those are averages across companies, not a quote. What moves your number:

Cost driverWhy it matters
ModelOpus 5.5 costs twice what Sonnet 5.5 does per token, and Fable 5.1 2.5x Opus
Session lengthEvery request resends the whole conversation, so long sessions cost more per message
BreaksAfter the cache expires, the next message reprocesses the full context at the full input price
Subagents and agent teamsEach one sends its own requests; the docs say teams in plan mode use about 7x the tokens
Parallel sessionsEach session is its own conversation with its own context
Thinking and effortThinking tokens are billed as output tokens

Most of these can be controlled. Our list of ways to cut Claude Code token usage goes through each one with the commands to use.

Subscription or API: which is cheaper?

There's no break-even number to quote, because Anthropic doesn't publish how many tokens a plan includes. But the trade-off is easy to reason about:

ChooseWhen
ProYou code with Claude Code a few hours a day and rarely hit the 5-hour limit
Max 5x or 20xYou hit Pro's limit often, run long sessions or several sessions at once
API keyUsage is light or irregular, you run Claude Code in CI or scripts, or your company pays for API usage
Team or EnterpriseYou need central billing, SSO, spend controls or admin analytics

A useful habit before deciding is to look at the dollar figure Claude Code computes for your sessions. On a subscription it isn't your bill; it's what the same tokens would cost on an API key. Our guide on API value versus what you pay explains how to read that number.

Ways you can pay more than you expect

A few settings quietly move you from a flat fee to per-token billing:

  1. An ANTHROPIC_API_KEY in your environment. If it's set, Claude Code asks once whether to use it, and if you approve, uses it instead of your subscription; in -p mode it always uses it. Anthropic's Pro and Max guide warns that this results in API charges. Run /status to see which login is active, and unset ANTHROPIC_API_KEY to fall back to the plan.
  2. Usage credits. On Pro and Max you can turn on usage credits to keep working past a limit at standard API rates. They're off until you enable them, and you can set a monthly spend cap in Settings > Usage.
  3. Fast mode. On subscription plans it's paid from usage credits only, even if you have plan usage left, according to the fast mode docs.
  4. Fable on Pro. It bills to usage credits, and Claude Code asks for consent before the first such request in an interactive session. In -p mode it doesn't ask.
  5. 1M context on older models. Sonnet 4.6 with 1M context needs usage credits on every plan. The current Sonnet 5.5 and Opus 5.5 run with 1M context on every plan without them.

How to check what you're spending

Inside Claude Code, run /usage (/cost and /stats are aliases). On an API key, its Session block shows tokens and an estimated dollar total for the session. On Pro and Max it shows plan usage bars and a breakdown of what used them, and the docs note that the session dollar figure isn't relevant for billing. For authoritative API billing, use the Claude Console usage page. For a running total at the bottom of every session, add cost.total_cost_usd to your Claude Code status line.

The Claude Code costs docs section Track your costs, explaining the Session block of /usage and showing sample totals
The costs docs: /usage shows the session's cost, computed locally at list price.

If you're on a Mac and run several sessions, Eddie, our notch app, shows tokens and API-equivalent cost per session, per project and per day, read from Claude Code's local logs with no API key. Like every tool that reads those logs, it shows list-price equivalents, not what a Pro or Max plan bills you. On any OS, our guide to checking Claude Code token usage covers every built-in option.

Eddie's open notch listing three agent sessions with their status, ports and cost today, and a Today total of 4.21 dollars and 1.9M tokens
On a Mac, in the interactive demo on editz.pro: cost per session and a Today total, read from Claude Code's local logs.

Prices in other countries

Anthropic's Pro plan article lists Pro at $20 a month in the US, with pricing in local currency where supported. Monthly prices vary by region, and some regions include tax in the displayed price while others add it at checkout. The Max prices above are for web subscriptions, and app store prices may differ. Anthropic also says it doesn't offer standing discounts, so a price that looks far below these numbers is either a regional price or a limited-time promotion.

For the exact price in India, the UK or anywhere else, open claude.ai/upgrade while signed in. API prices are in US dollars everywhere.

FAQ

How much does Claude Code cost per month?

Claude Code comes with every paid Claude plan: Pro is $20 a month billed monthly (or $17 a month on the $200 annual plan), Max 5x is $100 a month and Max 20x is $200 a month, before tax. With an API key you pay per token instead; Anthropic's docs put the enterprise average at about $150 to $250 per developer per month.

Is Claude Code free?

No. The Free Claude plan doesn't include Claude Code. You need Pro, Max, Team or Enterprise, or a Claude Console account that pays per token through an API key.

How much does Claude Code cost per token?

On an API key you pay the model's list price. As of October 3, 2026 that's $4 per million input tokens and $20 per million output tokens for Opus 5.5, $2 and $10 for Sonnet 5.5, and $1 and $5 for Haiku 4.5, with cache reads much cheaper. On Pro and Max you aren't billed per token at all unless you turn on usage credits.

Is the Max plan worth it over Pro for Claude Code?

If you hit Pro's 5-hour limit most days, yes: Max 5x gives five times Pro's usage per 5-hour session for $100, and Max 20x gives twenty times for $200. If you rarely hit the limit, Pro is enough. Check /usage for a week before you decide.

How much does Claude Code cost in India or the UK?

Anthropic lists Pro at $20 a month in the US and says local-currency prices vary by region, with tax included in some countries and added at checkout in others. Check claude.ai/upgrade or your app store for the exact price where you live.