March 27, 2026·15 min read

Claude Code Pricing: Every Plan and Price (2026)

Claude CodePricingGuidePlan Comparison

Claude Code pricing: every plan and price

Claude Code pricing starts at $20/month for the Pro plan and runs up to $200/month for Max 20x, with Team and Enterprise tiers above that and pay-per-token API billing for automation. The Claude Code price you pay depends on one thing: how heavily you use it. Anthropic runs two separate billing systems, subscriptions for interactive terminal use and API pay-per-token for programmatic use, and they do not overlap. Pick the wrong one and you either overpay or hit constant rate limits.

Here is the full Claude Code pricing at a glance:

Plan	Price	Best for
Pro	$20/month	Solo developers, a few hours of daily use
Max 5x	$100/month	Full-time devs who hit Pro limits daily
Max 20x	$200/month	All-day power users, parallel sessions
Team Premium	$125/user/month	Teams of 5+ needing central billing and admin
Enterprise	Custom (from ~$500/mo)	50+ devs, compliance, 1M context window
API	Pay-per-token	Scripts, CI/CD, product integrations

This guide covers every Claude Code plan and price available in 2026: what is included, the real usage limits, which tier fits which developer, and where the hidden costs appear. Whether you are deciding between Pro and Max or evaluating Claude Code against competitors like Cursor or Copilot, this breakdown gives you the numbers you need. If you are new to the tool itself, our introduction to Claude Code is a good starting point, and how to use Claude Code covers the workflow once you have picked a plan.

The two billing paths

Before looking at individual plans, understand the split:

Subscriptions (Pro / Max / Teams) - flat monthly fee, includes Claude Code for interactive terminal use. The UI at claude.ai, the desktop app, and the claude terminal command all draw from the same usage pool. Best for developers using Claude Code manually, in real time. This is the path most people mean when they ask about the Claude Code subscription price.

API pay-per-token - you pay per input/output token with no subscription. Required for automated, scripted, or CI/CD use. Subscriptions explicitly prohibit automated usage in their terms of service.

The typical developer using Claude Code in their terminal should be on a subscription. The developer embedding Claude in a product or running it in pipelines should be on the API. Most people fall clearly into one bucket.

Pro - $20/month

The entry point. Pro gives you Claude Code access plus the web interface, desktop app, and mobile apps, all drawing from a shared usage pool.

What you get:

Claude Code terminal access (Sonnet as the default model)
Approximately 10-40 Claude Code prompts per 5-hour rolling window
~44,000 tokens per 5-hour window (shared with chat usage)
Extended thinking mode
Unlimited projects
Priority over free-tier users during high demand

Rate limit mechanics: Pro uses a rolling 5-hour window. When you exhaust your token budget, you wait for the window to roll forward - typically 30 to 60 minutes. There is also a 7-day weekly ceiling that resets every seven days. Sustained heavy usage will hit both.

Who this fits: Developers using Claude Code for a few hours daily, exploratory use, smaller projects, or anyone starting out with Claude Code before committing to higher tiers. If you are not consistently hitting rate limits, Pro is the right plan. If you are, Max is the answer. For a deeper look at what $20 buys you, see our Claude Code Pro plan breakdown.

What Pro does not include: Opus model access, priority routing during peak demand (Max users get priority), and the usage headroom for extended sessions.

Max - $100/month (5x) and $200/month (20x)

Max is the subscription tier for developers who use Claude Code as a primary tool. It multiplies your Pro usage quota - 5x or 20x - and adds model access and routing advantages.

Max 5x - $100/month

What you get:

5x the Pro usage limits (~220,000 tokens per 5-hour window)
Full Opus model access (Claude's most capable model)
Priority traffic routing during peak demand
All Pro features included

Who this fits: Full-time developers running Claude Code throughout the workday. If you consistently hit Pro rate limits - especially during longer sessions or complex multi-file tasks - Max 5x resolves that. The Opus access alone is worth considering: it handles architectural reasoning and complex refactors at a noticeably higher capability level than Sonnet.

Max 20x - $200/month

What you get:

20x the Pro usage limits (~880,000 tokens per 5-hour window)
All Max 5x features included

Who this fits: Developers using Claude Code as their primary development environment for 6 to 8+ hours daily, or anyone running multiple concurrent Claude Code sessions. This is the tier for power users who cannot afford interruptions from rate limits mid-task.

Is Max worth it?

The honest answer depends on one thing: how often you hit Pro limits.

If you hit them rarely, Pro is the right call. If you hit them daily, Max 5x pays for itself almost immediately. The cost of context-switching away from a blocked tool, losing flow state, and waiting out a rate limit window adds up fast. If you are running Claude Code all day or running parallel sessions, Max 20x is the only tier that gives you the headroom for uninterrupted work. For the decision-tree version of this with the break-even numbers, see Claude Pro vs Claude Max.

For developers who use Claude Code daily, Max 5x at $100/month typically costs less than the equivalent API tokens would, which is what makes a flat subscription attractive for heavy interactive use.

Teams - from $25/user/month

Team plans add centralized billing, admin controls, and team management on top of the individual subscription features. Two seat types exist:

Team Standard - $25/user/month ($20 billed annually)

1.25x Pro usage per seat (Claude chat access, no Claude Code)
Centralized billing and admin dashboard
SSO integration
5-user minimum

Team Standard is for organizations that need centralized Claude access for non-engineering teams - marketing, product, support, operations - where Claude Code is not a requirement.

Team Premium - $125/user/month ($100 billed annually)

6.25x Pro usage per seat
Full Claude Code access
All Team Standard management features
5-user minimum

Team Premium is the right choice for engineering teams. At $125/seat it costs more than the individual Max 5x plan ($100), but includes the admin controls, centralized billing, and usage visibility that individual subscriptions do not provide. For teams where finance or security needs central oversight over AI tool spend, the premium is typically justified.

Team plan note: Seat sharing is not permitted. Anthropic's terms prohibit sharing credentials, and accounts flagged for this are terminated. Each developer needs their own seat.

Enterprise - custom pricing

Enterprise is for organizations with 50+ developers or specific security, compliance, or procurement requirements.

What Enterprise adds over Team Premium:

1M token context window (vs. 200K on consumer plans)
SCIM provisioning for automated user management
Audit logging for compliance and security reviews
Compliance APIs for SOC 2, HIPAA, and similar requirements
Dedicated support
Custom usage guarantees (SLA-backed uptime and capacity)
Tailored contract terms

Pricing: Custom. Reported figures from 2026 range from $500 to $1,000/month for small deployments (10-25 developers) to $5,000-$15,000+/month for 100+ person engineering organizations. Enterprise pricing typically involves both a per-seat fee and a usage tier.

When Enterprise is the right choice: When your organization has compliance requirements that consumer plans do not satisfy, when you need SCIM for automated onboarding/offboarding, when a 200K context window is a bottleneck (1M is a material difference for large codebases), or when legal requires custom contract terms. Our Claude Enterprise pricing guide breaks down the $20-per-seat-plus-API-usage structure and when it beats Teams Premium.

API pay-per-token - the programmatic path

Subscriptions cover interactive use. For automation, integrations, CI/CD pipelines, or building products on top of Claude, you need API billing.

Current API rates (2026):

Model	Input (per 1M tokens)	Output (per 1M tokens)
Claude Opus 4.8	$15	$75
Claude Sonnet 4.6	$3	$15
Claude Haiku 4.5	$1	$5

API tier rate limits scale with cumulative spend:

Tier	RPM	TPM (input)	How to reach
Tier 1	50	40,000	$5 deposit
Tier 2	1,000	80,000	$40 spend
Tier 3	2,000	160,000	$200 spend
Tier 4	4,000	400,000	$400 deposit

The critical rule: subscriptions and API access are entirely separate. A Max subscription does not grant API tokens, and API keys do not provide subscription features. If you are automating Claude Code in any way - scripts, hooks that call the API, CI/CD - you need an API key with a funded balance, separate from your subscription. For the step-by-step on generating and configuring one, see our Claude Code API keys guide.

Claude Code vs. competitor pricing

Understanding how Claude Code pricing compares to alternatives helps frame the value proposition:

Tool	Monthly cost	What you get
Claude Code Pro	$20	Terminal AI agent, Sonnet model, ~44K tokens/5hr
Claude Code Max 5x	$100	All Pro + Opus access, 5x limits
Cursor Pro	$20	IDE-integrated AI, 500 fast completions/mo
Cursor Business	$40	All Pro + admin, centralized billing
GitHub Copilot Individual	$10	Code completions, chat, limited agent
GitHub Copilot Business	$19	All Individual + org management

The price-to-capability ratio differs fundamentally. Cursor and Copilot are IDE plugins that enhance your editor. Claude Code is an autonomous agent that reads, writes, debugs, deploys, and manages your codebase from the terminal. The $20 Pro tier competes with Cursor Pro on price but operates at a different capability level - Claude Code can run multi-step tasks, spawn subagents, and maintain persistent memory across sessions.

For a detailed feature comparison, see our Claude Code vs Cursor analysis.

Choosing the right plan

Situation	Recommended plan
Testing Claude Code or occasional use	Pro ($20/mo)
Daily use, hitting Pro limits regularly	Max 5x ($100/mo)
Claude Code is your primary dev tool, full-day use	Max 20x ($200/mo)
5+ developers, need centralized billing/admin	Team Premium ($125/seat/mo)
50+ developers, compliance requirements	Enterprise (custom)
Automation, scripts, CI/CD pipelines	API pay-per-token

Cost optimization strategies

Whichever plan you are on, these strategies reduce wasted tokens and increase the productive output per dollar.

Use the right model for each task

Opus costs roughly 5x as much per token as Sonnet, and Sonnet about 3x as much as Haiku. Use Haiku for search, formatting, and simple queries. Use Sonnet for standard coding tasks. Reserve Opus for complex multi-file reasoning and architecture decisions. This alone can cut your effective token burn by 30-50%.

Compact aggressively

Running /compact during long sessions summarizes earlier context and frees up token budget. The conversation thread stays alive, but the context window reclaims space. On Pro, compacting every 20-30 minutes during intensive work can extend a session significantly.

Clear between tasks

Context from one task pollutes the next. Clearing between major tasks is not just a quality practice - it is a cost practice. A fresh context window means Claude does not send irrelevant history with every request, which directly reduces token consumption.

Configure CLAUDE.md and memory

A well-written CLAUDE.md eliminates the "explain the project" tokens you spend at the start of every session. A memory system eliminates the "remind me what we decided" tokens mid-session. These configuration files are a one-time investment that pays back on every session.

Read specific lines, not entire files

Every file read counts against your token budget. Reading 50 lines of a 500-line file uses 10% of the tokens. Use Grep to find the relevant section first, then read just those lines.

For more strategies on working within rate limits, see our rate limit fix guide.

The hidden cost: setup and configuration

Here is where most developers underestimate the real cost of Claude Code: the tool itself is just the subscription. Getting useful output from it requires configuration - CLAUDE.md project files, persistent memory systems, custom skills, hooks, and agent workflows.

Out of the box, Claude Code is capable but generic. It does not know your codebase conventions, does not remember decisions from last week, and does not follow your deployment process. Each of those gaps requires time to close.

The configuration investment is real:

Configuration layer	Time to build from scratch
CLAUDE.md project file	~10 minutes
Persistent memory system	~15-30 minutes
5-10 custom skills	1-2 hours
Hooks and quality gates	~30 minutes
Agent workflows	2-4 hours

Once configured, Claude Code's output quality improves dramatically and the value-per-dollar of any subscription tier increases. A well-configured Pro subscription outperforms a default Max subscription for most tasks, because configuration determines output quality more than raw token limits.

For a walkthrough of building this from scratch, the complete Claude Code setup guide covers every layer.

Rate limits in practice

Understanding the rate limit mechanics helps you work around them rather than getting blocked:

The 5-hour rolling window is the primary limit. When you exhaust your token budget, you cannot resume until the window rolls forward. The window does not reset at a fixed time - it rolls from when you started using tokens.

Strategies that help:

Run /compact regularly in long Claude Code sessions - this summarizes the conversation history and dramatically reduces token consumption per request
Start new sessions at natural task boundaries instead of continuing one long session
On Pro, schedule heavy Claude Code work in bursts: intense session, natural pause, fresh session. This works with the rolling window rather than against it

What burns tokens fast in Claude Code:

Large codebase reads (Claude reads entire files by default)
Long conversation histories being sent with each request
Back-and-forth iteration without compacting
Opus model usage vs. Sonnet (Opus is significantly more token-intensive)

How Claudify fits in

Claudify does not replace any Anthropic plan - it enhances whichever plan you are on. It is a pre-built configuration stack for Claude Code: 1,727 skills across 31 categories, persistent memory architecture, 9 specialist agents, and 21 slash commands. One install, immediate productivity gain.

The economics work at every tier. On Pro, the right configuration means you extract more value per token - fewer wasted requests from Claude guessing at your conventions, more precise task execution, less back-and-forth. On Max, the configuration headroom means your 20x token budget actually goes toward productive work rather than context reconstruction.

Claudify is a one-time $49 purchase - no subscription, no per-seat pricing. It works across all Claude Code plans including Team and Enterprise deployments. See what's included.

Frequently asked questions

How much does Claude Code cost?

Claude Code pricing has four subscription tiers plus API billing. The entry price is $20/month for Pro. Max is $100/month for 5x usage or $200/month for 20x usage. Team Premium is $125 per user per month with a 5-user minimum, and Enterprise is custom (reported figures start around $500/month for small deployments). For automated or programmatic use, you pay per token on the API instead of a flat monthly Claude Code price. The right plan is whichever one stops you hitting rate limits without paying for headroom you do not use.

Is Claude Code free to use?

Claude Code requires at minimum a Pro subscription ($20/month). There is no permanent free tier for Claude Code specifically, though Anthropic occasionally offers limited free access to the Claude chat interface. The Pro plan is the entry point that includes terminal access, and it provides enough usage for developers working a few hours daily. If you want to evaluate Claude Code before committing, the Pro plan gives you a full month to test whether it fits your workflow.

What is the best Claude Code plan for a solo developer?

Start with Pro at $20/month. Use it for a full week of regular development work. If you hit rate limits more than twice in that week, upgrade to Max 5x. Most solo developers working 4-6 hours daily find Max 5x ($100/month) to be the sweet spot - it provides enough headroom for uninterrupted work plus access to the Opus model for complex reasoning tasks. Max 20x is only necessary if you are running Claude Code for 8+ hours daily or using multiple concurrent sessions.

Can I use one Claude Code subscription for my whole team?

No. Each developer needs their own subscription or team seat. Anthropic's terms explicitly prohibit credential sharing, and accounts flagged for multi-user access are terminated. For teams of 5 or more developers, Team Premium ($125/seat/month) provides centralized billing, admin controls, and usage visibility. For smaller teams, individual Max subscriptions are simpler and slightly cheaper per seat.

Is the Claude Code Max price worth it over Pro?

It depends entirely on how often you hit Pro's rate limits. If you rarely hit them, the $20 Pro price is the right call and Max is wasted money. If you hit them daily, Max 5x at $100/month usually pays for itself, because the cost of losing flow state and waiting out a 5-hour window adds up fast. Max also unlocks Opus and priority routing. Run Pro for a full week first, then upgrade only if rate limits keep interrupting your work. See our Claude Pro vs Claude Max breakdown for the break-even maths.

Does the Claude Code price include API access?

No. Subscriptions and API billing are entirely separate. A Pro, Max, or Team subscription covers interactive terminal use only and grants zero API tokens. If you automate Claude Code through scripts, hooks, or CI/CD, you need a funded API key billed per token on top of (or instead of) your subscription. This separation is the single most common source of confusion in Claude Code pricing.

Summary

Claude Code pricing in 2026 has four clear subscription tiers plus pay-per-token API access:

Pro ($20/mo) - entry point, good for several hours of daily use
Max 5x ($100/mo) - professional developers hitting Pro limits daily
Max 20x ($200/mo) - heavy use, all-day sessions, power users
Team Premium ($125/seat/mo) - 5+ developers with centralized management needs
Enterprise (custom) - 50+ developers, compliance requirements, 1M context window
API (pay-per-token) - automation, scripts, CI/CD, product integrations

The right plan is the one that eliminates rate limit interruptions without over-spending on headroom you do not use. Start at Pro. Upgrade to Max when you consistently hit limits. Add Teams when your organization needs centralized oversight.

The bigger productivity lever - at any price point - is configuration. A well-configured Claude Code setup changes the return you get from every token, regardless of which tier you are on. For the best practices that maximize your plan's value, start with CLAUDE.md, memory, and context hygiene.