Claude Code Pricing 2026: Plans, Limits, and Hidden Costs

Confused by Claude Code pricing in 2026? Plans range from $0 to $200/month for individuals and $20 to $125/seat for teams, with weekly usage limits that can automatically switch a subscription to pay-per-token mode without warning.

Claude Code pricing at a glance

Claude Code ships inside Anthropic's paid Claude plans. The same subscription covers Claude on web, desktop, and mobile, as well as the Claude Code command-line tool.

Here's how the pricing tiers compare:

Pricing correct as of June 2026. See Anthropic's pricing page for the latest rates.

How Claude Code pricing works

Anthropic bundles Claude Code into subscription plans alongside Claude.ai chat and Claude Cowork. A Pro subscription covers all three under one price, with usage shared across them.

That bundling has two implications worth understanding before you pick a plan.

First, your Claude Code usage counts against the same limits as your Claude chat usage.

If you spend the morning chatting with Claude in the browser and then switch to Claude Code in your terminal, both eat into the same five-hour session and weekly caps.

Second, the per-token cost of Claude Code through a subscription is effectively much lower than the same usage on the API, but only until you hit your plan's limit.

After that, you either wait for the reset or pay for additional usage at API rates.

Free plan and why Claude Code isn't included

The Free plan gives you access to Claude on web, iOS, Android, and desktop, but does not include Claude Code. If you want to try Claude Code without paying upfront, the lowest-cost path is one month of Pro at $20.

What the Free plan does include:

• Chat on web, iOS, Android, and desktop.
• Generate code and visualize data inside chat.
• Web search and memory across conversations.
• Create files and execute code (with limits).
• Connect Slack and Google Workspace services.
• Extended thinking for complex work.

What's missing for developers:

• No Claude Code (terminal access requires Pro or higher).
• No Claude Cowork (the desktop file and task agent).
• Limited usage that kicks in fast on coding-heavy work.
• No access to all models (Opus 4.8 and Fable 5 are gated to paid tiers).

The Free plan works for evaluating Claude's writing and chat capabilities. Claude Code itself requires a paid plan to test meaningfully.

See Anthropic's Free plan page for current feature parity and limits.

Pro plan for daily individual coding

The Pro plan is the entry point for using Claude Code. At $20/month on a monthly plan or $17/month on an annual plan ($200 upfront), it covers a single user across all three Claude products.

What you get on Pro:

• Claude Code in your terminal: with shared usage limits across Claude products.
• Claude on web, desktop, and mobile: with the full chat experience.
• Claude Cowork: desktop agent for file and task automation.
• Access to Claude's full model lineup: Opus 4.8 for ambitious agentic coding, Sonnet 4.6 for the best mix of capability and price, and Haiku 4.5 for fast, low-cost runs. Fable 5 access ships in promotional mode on consumer plans (prior to June 12, 2026).
• Claude for Microsoft 365 and Claude for Microsoft Outlook: two distinct integrations covering Office documents and email respectively.
• Unlimited projects: to organize chats, code, and knowledge bases.

Usage limits reset weekly and across five-hour sessions, with the exact token count Anthropic permits varying by model selection.

Most light-to-moderate Claude Code users (a few hours per day, mainly Sonnet) stay within Pro limits. Heavy Opus users on large repositories typically hit caps and need Max.

See Anthropic's Pro plan page for current pricing and feature details.

Max plan for heavy individual users

The Max plan is for individual developers who consistently hit the Pro limits while coding. It comes in two tiers.

Max 5x ($100/month):

• 5x Pro capacity per five-hour session.
• Higher output limits for all tasks.
• Early access to advanced Claude features.
• Priority access during high-traffic periods.

Max 20x ($200/month):

• 20x Pro capacity per five-hour session.
• Everything in Max 5x with substantially higher caps.

For Claude Code specifically, the jump from Pro to Max 5x makes sense if you regularly hit usage limits during the workday, especially when running multi-step coding tasks on large repositories with Opus.

Max 20x is for power users who treat Claude Code as their primary coding tool and run it for several hours a day.

The math: $200/month is roughly $6.67/day, which beats the API rate for sustained Opus usage.

See Anthropic's Max plan page for current pricing and capacity-per-session figures.

Team plan for collaborative teams

The Team plan is the entry-level option for organizations that want Claude Code (plus Claude chat) for multiple people under centralized billing and admin controls. It requires a minimum of 5 members and tops out at 150 seats.

Team plans come in two seat types you can mix within the same organization:

Standard seats:

• $25/seat/month on a monthly plan or $20/seat/month on an annual plan.
• More usage per session than Pro.
• Single sign-on (SSO) and domain capture.
• Admin tools and billing management.
• Claude Code is not included on Standard seats (Premium only).

Premium seats:

• $125/seat/month on a monthly plan or $100/seat/month on an annual plan.
• 5x more usage per session than Standard seats.
• Same governance features as Standard, with higher caps.
• Includes Claude Code access alongside the other Claude apps.

Premium seats are the only Team option that bundles Claude Code, while both seat types include Claude Cowork, enterprise search, and connectors to Microsoft 365, Slack, GitHub, Google Drive, and Gmail.

Team plans also support purchasing usage credits that let members continue working after hitting their seat's included limits.

That's the lever to pull if some users consistently exceed their Standard seat capacity without a valid justification for a Premium upgrade. Organizations with more than 150 seats need to upgrade to Enterprise.

See Anthropic's Team plan page for current per-seat pricing and feature parity.

Enterprise plan for large organizations

The Enterprise plan is built for larger organizations that need everything in Team plus deeper governance, advanced identity management, and tailored procurement terms. Pricing starts at $20/seat with usage billed at API rates.

What Enterprise adds over Team:

• Role-based access control: with fine-grained permissioning.
• SCIM provisioning: for automated user lifecycle management.
• Audit logs and Compliance API: for observability and monitoring.
• Custom data retention controls: for regulated industries.
• Network-level access control: and IP allowlisting.
• HIPAA-ready offering: available for healthcare deployments.
• Claude Security (beta): for security teams investigating threats.

Enterprise pricing separates access from usage: $20/seat covers platform access, while usage is billed separately at API rates.

That model favors organizations with variable usage across their workforce, since light users don't subsidize heavy ones.

For larger contracts, Enterprise also supports invoicing, net terms, AWS Marketplace, and non-standard MSAs through Anthropic's sales team.

See Anthropic's Enterprise plan page for current pricing and feature details.

Claude API pay-as-you-go pricing for developers

Claude Code also accepts API credits as authentication and billing usage at standard per-token rates.

This makes sense for developers running intensive sprints or for teams already using Claude through Anthropic's API.

Current API pricing for the most-used Claude models:

Note: As of June 12, 2026, Anthropic removed access to Fable 5 and Mythos 5 for all users in response to a US export control directive that bars access by foreign nationals. Other Claude models are unaffected.

MTok = million tokens. Prompt caching can dramatically reduce costs for repeated prompts, especially in codebases where the same context is loaded repeatedly.

Important catch: If you have an ANTHROPIC_API_KEY environment variable set, Claude Code authenticates with the API key, charging your API account at pay-per-token rates.

Many developers get surprised by API bills this way.

To stick with your subscription, run "claude logout" and re-authenticate using only your Pro or Max credentials. To allow API credit fallback when subscription limits hit, leave the environment variable set.

See Anthropic's API pricing page for the latest per-token rates and prompt caching details.

Hidden costs and usage limits to watch

Claude Code's headline pricing is straightforward, but a handful of usage patterns drive up real-world costs. These are the ones worth knowing before you sign up.

Usage limits reset weekly, not monthly

Both Pro and Max plans have weekly usage caps in addition to per-session limits. If you blow through your week's allocation by Wednesday, you wait until your session anniversary to keep coding on the subscription.

Sessions are 5 hours each

Each Claude session lasts five hours, during which usage accumulates against your session cap. Multiple sessions can be active in a week, but each five-hour window has its own ceiling.

Opus burns through limits faster than Sonnet

Opus is roughly 1.7x more expensive than Sonnet on the API ($5 vs $3 input, $25 vs $15 output), and the same ratio applies to how quickly it consumes subscription capacity.

Default to Sonnet 4.6 for most coding tasks and reserve Opus for the hard ones.

The ANTHROPIC_API_KEY trap

If you set an API key as an environment variable, Claude Code still uses API credits even when you have a subscription. Check your billing source with the /status command in Claude Code.

API web search and code execution have separate costs

When using Claude Code via the API, web search costs $10 per 1,000 searches and code execution costs $0.05 per container-hour after the 50 free hours per day. Subscription plans don't have these line items.

Team plan usage credits add up

Team plans can enable usage credits for members who exceed their seat limits. These bill at API rates and aren't capped by default, so set spend controls early.

Which Claude Code plan should you choose

The right plan depends on whether you're an individual developer, part of a team, or running Claude Code at organizational scale. Here's how to map the decision.

Choose Pro ($17-$20/month) if you:

• Are an individual developer using Claude Code for daily coding work.
• Want the cheapest legitimate way to access Claude Code.
• Don't hit Pro's usage limits more than once or twice a week.

Choose Max 5x ($100/month) if you:

• Regularly hit Pro limits during the workday.
• Run multi-step coding tasks on larger repositories.
• Want priority access during high-traffic periods.

Choose Max 20x ($200/month) if you:

• Treat Claude Code as your primary coding tool.
• Run Opus heavily for hours every day.
• Want the highest individual capacity Anthropic offers.

Choose Team ($20-$125/seat/month) if you:

• Have 5 or more developers who need Claude Code access (requires Team Premium seats at $100-$125/seat/month, since Standard seats don't include Claude Code).
• Need SSO, domain capture, and centralized billing.
• Want a mix of Standard and Premium seats for different user types.

Choose Enterprise ($20/seat + usage) if you:

• Have more than 150 seats or need RBAC, SCIM, and audit logs.
• Operate in a regulated industry (financial services, healthcare, government).
• Need custom MSAs, invoicing, or AWS Marketplace procurement.

Choose API pay-as-you-go if you:

• Build agents, tools, or integrations that programmatically call Claude.
• Have spiky usage that doesn't fit a flat monthly subscription.
• Already manage Anthropic API credits for other workloads.

When AI work moves beyond CLI coding

Claude Code is exceptional at terminal-based coding for individual developers and engineering teams. It's the right tool when you're generating code, refactoring repositories, or running coding tasks directly from the command line.

The picture changes when your AI work moves beyond code generation into building internal apps that business teams use, with a UI, role-based permissions, and audit trails on every action.

That's a different category from CLI coding, and where governed AI app platforms come in.

Superblocks is one of these platforms, with Clark AI (its app-building agent) powered by Claude Opus. Business teams can build apps from natural-language prompts, while IT configures the guardrails (RBAC, SSO, audit logs, BYO inference) once and applies them to every app the platform produces.

For teams who want Claude Opus paired with governance for business-team app building, our introduction to Superblocks walks through the setup.

Or book a demo to test it with your own stack.

Frequently asked questions

Is Claude Code free?

No. Claude Code requires a Pro, Max, Team Premium, or Enterprise subscription starting at $20/month on a monthly plan (or $17/month with an annual plan), or you can pay per token through the Claude API.

How much does Claude Code cost per month?

Claude Code costs $20/month on Pro monthly (or $17/month annual), $100/month on Max 5x, $200/month on Max 20x, or $100-$125/seat/month on Team Premium seats (Standard seats don't include Claude Code), plus optional API credits if you exceed limits.

Does Claude Code count against my Claude usage limits?

Yes, Claude Code shares usage limits with the Claude chat interface across all paid plans, meaning every coding session counts against the same five-hour session and weekly usage caps that apply to web, desktop, and mobile chats.

What's the difference between Pro and Max for Claude Code?

The difference between Pro and Max for Claude Code is usage capacity: Pro gives you standard access at $17-20/month; Max 5x gives 5x more capacity per session for $100/month; and Max 20x gives 20x more capacity for $200/month.

Can I use Claude Code with API credits instead of a subscription?

Yes, Claude Code accepts API credits directly through a Claude Console account, charging standard API rates per token ($3 to $5 per million input tokens depending on the model).

Is Claude Code worth it compared to GitHub Copilot or Cursor?

Claude Code is worth it for developers who prefer terminal-based workflows to IDE plugins, want access to Claude Opus for complex coding tasks, and already use Claude.ai, since one Pro or Max subscription covers both products.