Claude Code Pricing: Is It Free and Which Plan Fits?

Claude Code pricing confuses people because Claude Code has no price of its own. It comes bundled with every paid Claude plan, or you pay per token through the API. It isn't on the free plan.

So the real question is which plan matches how much you'll use it.

I pay for Claude Code myself and run skills, content audits and Obsidian brains through it most days, so I've learned where the limits bite. Every figure below comes from Anthropic's official pages as of September 2026, linked in the sources.

Is Claude Code free?#

No. The comparison table on Anthropic's pricing page lists Claude Code as unavailable on the Free plan and included on Pro, Max, Team and Enterprise.

According to the Claude Code quickstart, you can sign in with one of four account types:

• A Claude subscription: Pro, Max, Team or Enterprise
• A Claude Console account with prepaid API credits
• A cloud provider: Amazon Bedrock, Google Cloud or Microsoft Foundry
• A self-hosted gateway your organization runs

If you want to try it without a subscription, a Console account with a small prepaid balance is the closest thing to a free trial. You'll pay per token, so keep sessions short.

Claude Code plans compared#

A few details that matter in practice:

• Every plan shares one pool. Chat, Cowork, Claude Code and the desktop apps all draw from the same usage allowance.
• Limits reset on a rolling five-hour window. Paid plans also have weekly limits on top.
• Context window is 200k tokens on consumer and Team plans, and 500k on Enterprise's default model.
• Usage credits let paid plans keep working past a limit at standard API rates.

Anthropic's page lists Max as "from $100". Composio's June 2026 comparison puts the 20x tier at $200 a month, which matches what most heavy users report paying.

How much does Claude Code cost on the API?#

If you sign in with a Console account, you pay per million tokens (MTok) at these rates:

Prompt caching cuts the cost of repeated context, and Claude Code uses it automatically.

How much does that add up to? Anthropic's cost documentation reports that across enterprise deployments the average is around $13 per developer per active day and $150 to $250 per developer per month, with most developers staying under $30 per active day. Your numbers will differ with model choice, codebase size and how many sessions you run at once.

One trap worth knowing: if an ANTHROPIC_API_KEY environment variable is set in your shell, Claude Code asks to use that key, and approving it bills the API instead of your subscription.

Which Claude Code plan should you choose?#

Here's how I'd decide:

• Trying it for a week: Console with prepaid credits, or one month of Pro.
• A few focused sessions a day: Pro. Annual billing brings it to $17 a month.
• Claude Code is your main tool all day: Max 5x. Most people who hit Pro's limits daily are fine here.
• Running several agents in parallel or long autonomous jobs: Max 20x, or the API with spend limits.
• A team of two or more: Team, so you get central billing, admin controls and no model training on your content by default.

If you're comparing against OpenAI's agent before paying, my Claude Code vs Codex comparison covers how far each $20 plan stretches.

How do you make your Claude Code usage last longer?#

Most usage problems come from context, not from the number of messages. Claude Code sends your whole conversation with every request, so a long-running session costs more per question as it grows. These habits, all from Anthropic's cost guide, make the biggest difference:

1. Run /clear between unrelated tasks. Stale context is paid for on every message.
2. Default to Sonnet. Save Opus for architecture and hard reasoning, and use Haiku for simple subagents.
3. Disable MCP servers you aren't using. Run /mcp and switch off the idle ones; a CLI such as gh is often lighter.
4. Keep CLAUDE.md under 200 lines. Move long, task-specific instructions into skills that load only when needed. My guide to writing CLAUDE.md shows how.
5. Use plan mode for big changes. Press Shift+Tab to plan first, so a wrong direction doesn't burn a session.
6. Send verbose work to subagents. Test runs and log reading stay in the subagent's context and only a summary comes back.
7. Check /usage weekly. It shows plan usage bars and which skills, subagents and MCP servers consume the most.

Is Claude Code worth the price?#

For me, yes, but only because I use it on real work every day. What you're paying for isn't autocomplete; it's an agent that can take a task, work through a repository or a folder of notes, and come back with something reviewable.

If you're new to it, start with Pro for a month, follow a structured first week like the one in how to use Claude Code, and only upgrade when you hit limits on work that's actually paying off. And if you'd rather learn it live on your own projects, my Claude Code course covers cost control in week four.