TL;DR

No. Claude Code isn't part of Anthropic's free Claude plan. Anthropic's pricing page lists Claude Code as "No" for Free and "Yes" for Pro, Max 5x, Max 20x, Team and Enterprise.

The installer itself costs nothing, so "free to download" is true. Using it needs one of these:

• A Claude subscription. Pro is $20 a month billed monthly, or $17 a month on the annual plan. Max starts at $100 a month. Claude Code shares the plan's usage limits with Claude chat.
• Pay-as-you-go API billing through a Claude Console account. You pay per token at API rates, with no monthly fee.
• A cloud provider such as Amazon Bedrock, Google Cloud's Agent Platform or Microsoft Foundry, billed per token to your cloud account.

Prices here were read from Anthropic's pages on 26 October 2026. Anthropic says prices and plans can change, so check the pricing page before you buy.

The problem

The question usually comes up right after installing it. The install command works, claude starts, and then it asks you to sign in. A free claude.ai account can't use it, and the pricing page has several plans, a Console, and per-token rates for half a dozen models. It isn't obvious which route costs less for the way you work.

The honest answer depends on how much you use it. A few short sessions a week and a full workday of agent runs cost very different amounts, and the two billing systems behave differently when you run out.

What Anthropic's pricing says about Claude Code

From the pricing page and Anthropic's help article on using Claude Code with Pro or Max:

A few details matter more than the headline price:

• The limits are shared. Anthropic's pricing FAQ says activity across Claude on web, desktop, mobile and Claude Code all draws from the same pool. A long chat in the browser leaves less for the terminal.
• There's no fixed message count. How far a plan goes depends on conversation length, the model you choose and the features you use.
• Max's included API credits don't cover Claude Code. The help article says the monthly API credits included with Max can't be used in Claude Code or the Claude apps.

Route 1: A Pro or Max subscription

This route suits an individual developer who wants a fixed monthly bill. You sign in with /login, pick your claude.ai account, and usage counts against the plan.

What you get:

• A predictable monthly bill.
• A five-hour session limit and a weekly limit. Anthropic's help center says both apply on Pro and Max.
• Usage credits as an opt-in extra. If you turn them on, work continues past a limit at standard API rates until your spend limit.

What to watch:

• A stray API key overrides the subscription. Anthropic's help article warns that if ANTHROPIC_API_KEY is set in your environment, Claude Code uses it instead of your plan, and you get API charges. Run echo $ANTHROPIC_API_KEY before you start a long session.
• Heavy days hit the ceiling. Long sessions, big context and subagents all burn through the window faster. See when Claude Code usage resets for how the windows work.

Route 2: Pay-as-you-go through the Claude Console

With a Console account you pay per token. There's no monthly fee and no five-hour window, but there's also no ceiling unless you set one.

Anthropic's costs page gives a sense of scale. Across enterprise deployments, it reports an average of around $13 per developer per active day and $150 to $250 per developer per month, with 90% of users staying below $30 per active day. Your numbers will differ. A light user can spend much less than a Pro subscription, and a heavy user can spend far more than Max.

Per-token prices vary by model. The pricing page lists, for example, Sonnet 5.5 at $2 per million input tokens and $10 per million output tokens, and Opus 5.5 at $4 input and $20 output. Thinking tokens bill as output tokens, so effort level changes the bill too.

Cost controls on this route:

• Set a workspace spend limit in the Console.
• Run /usage to see the session's estimated cost. Claude Code computes it locally at list price, so the Console's Usage page is the authoritative bill.
• Use --max-budget-usd on scripted runs to cap one job.

Route 3: A cloud provider or gateway

If your company already pays for Amazon Bedrock, Google Cloud's Agent Platform or Microsoft Foundry, Claude Code can bill there per token. Spend controls live in that cloud's billing console, and Anthropic's analytics don't see the usage. This route is mostly for organizations, and it's the reason some teams can use Claude Code without anyone holding a personal subscription.

Which route fits

• A few sessions a week: pay-as-you-go may cost less than Pro. Watch the Console for the first month and compare.
• Daily use, one person: Pro is the usual starting point. Move to Max if you keep hitting the five-hour window.
• All-day agent runs: Max 5x or 20x, or API billing with a spend limit if you'd rather pay for exactly what you use.
• A company account: Team, Enterprise or your cloud provider, depending on what your organization already has.

Whatever you pick, the habits that save usage are the same. Clear context between unrelated tasks with /clear, keep CLAUDE.md short, and match the model to the job. Anthropic's costs page recommends all three.

Where Happier fits

Happier doesn't make Claude Code free and doesn't add allowance to any plan. Your subscription or API key pays for the model either way.

If you want to keep an eye on what's left of a plan, or keep a session going when one of your own accounts runs out, from your phone, Happier is an open-source app that lets you run and orchestrate Claude Code, Codex and other agents cross-device, from your phone, desktop or the web.

The relevant parts:

• Connected services store the subscriptions and API keys you already hold, encrypted, and reuse them across your computers. Subscription auth and API-key auth stay separate service types, so a Claude subscription and an Anthropic API key never get mixed up.
• Quota meters show what remains on a Claude subscription, a Codex subscription or Gemini, where the provider reports it. Other agents have no meter.
• Account pools hold accounts you own, such as a personal and a work subscription. When one reaches its limit, Happier can switch the session to the other and continue.

Connected services and their quota meters are experimental and off by default in Happier 0.2.17. Turn them on under Settings → Features. Setup is covered in Run Happier from your phone. Happier is on GitHub, you can download it here, and there's a Discord.

Common mistakes

• Assuming the free plan includes a trial of Claude Code. The pricing table lists it as unavailable on Free.
• Leaving ANTHROPIC_API_KEY exported while signed in with Pro. The key wins and the usage is billed per token.
• Expecting Max's API credits to pay for Claude Code. The help article says they can't.
• Comparing a subscription to API spend after one day. Usage swings a lot from day to day. Compare over a few weeks.

FAQ

Is Claude Code free to install?

Yes. The CLI, the IDE extensions and the desktop app download without charge. You need a paid plan or an API account to use them.

Do you need Claude Pro for Claude Code?

You need a paid plan or another billing route. Pro is the entry plan that includes it. A Console account with API billing, or a cloud provider, also works without any subscription.

Can you use Claude Code with a free claude.ai account?

No. The pricing table marks Claude Code as unavailable on the Free plan.

What happens when a Pro plan runs out?

Claude Code blocks requests until the reset time in the message, unless usage credits are on. What to do when you hit the Claude Code usage limit covers the options.

Does Claude Code cost more than Claude chat?

On a subscription it's the same plan and the same limits. Coding turns tend to use more of the allowance, because each turn carries file contents and tool results.

Related guides

• When does Claude Code usage reset?
• How to check Claude Code usage
• What to do when you hit the Claude Code usage limit
• How to use your Codex subscription on multiple computers
• How to log in to Claude Code on a remote server over SSH