Claude Code pricing: plans, usage costs and real examples

Claude Code pricing starts at $20 a month. Claude Code is included in Claude Pro ($20/month billed monthly, or $17/month billed annually) and in every higher paid plan: Max, Team and Enterprise. It isn't on the Free plan. If you'd rather pay per use, you can run Claude Code on a Claude Console account and pay per token at API rates.

By Gary, Greenlight My App · Checked on Oct 21, 2026; based on Anthropic's public sites (claude.com/pricing (external), Claude Code cost docs (external)) · Drafted with AI tools and reviewed before publishing.

I use coding agents every day while building and checking AI-made apps, and the question I hear most is "what will this actually cost me?" This guide answers that for Claude Code specifically: which plan includes it, how pay-as-you-go billing works, and what a few typical days cost. For per-model API rates, batch discounts and tool fees, see our full Claude API pricing breakdown; this page doesn't repeat it.

How much does Claude Code cost?

It costs $20 a month on the cheapest plan that includes it, or whatever your tokens add up to on pay-as-you-go. Here are the two ways to pay.

1. A Claude subscription. Claude Code is listed as included in Pro, Max 5x, Max 20x, Team and Enterprise on Anthropic's pricing page (external). You pay a flat monthly price and work within that plan's usage limits.
2. A Claude Console account (API billing). Per Anthropic's cost docs (external), "Claude Code charges by API token consumption." You pay for every token sent and received, with no monthly fee.

Which Claude plans include Claude Code?

Prices as of Oct 21, 2026. Check the official page before you rely on them: claude.com/pricing (external).

Two details from Anthropic's pricing FAQ matter for coders:

• One shared pool. "Your activity across Claude on web, desktop, mobile, and Claude Code all draws from the same pool." A long chat session leaves less room for coding, and the reverse.
• What happens at the limit. You can wait for the reset, move to a higher plan, or, on paid plans, turn on usage credits "to keep working at standard API rates." Usage credits are optional and you can set a monthly spend limit.

How does pay-as-you-go Claude Code billing work?

You pay per million tokens (MTok) at the model's API rate. Every request carries your prompt, the files Claude reads, tool results and the conversation so far (input), plus Claude's reply and its thinking (output). Thinking tokens are billed as output, per the cost docs (external).

Prompt caching lowers the cost of re-sending the same context, and Claude Code uses it automatically. Rates per model live in our Claude API pricing guide. For the examples below I use two list rates from claude.com/pricing (external), checked Oct 21, 2026:

Prices as of Oct 21, 2026. Check the official page before you rely on them.

What does Anthropic say a typical developer spends?

Anthropic's own figure: "Across enterprise deployments, the average cost is around $13 per developer per active day and $150-250 per developer per month, with costs remaining below $30 per active day for 90% of users." That's an average across companies, not a quote for you. Anthropic suggests starting with a small pilot and tracking real usage before you budget.

Worked examples: what a Claude Code day costs

These are simple, uncached list-price estimates so you can see the math. Real bills are usually lower for repeated context because of caching, and higher if a session runs all day.

Example 1: a focused bug fix on Sonnet 5.5. Claude reads a few files and iterates five or six times. Say the session totals 1,500,000 input tokens and 100,000 output tokens.

• Input: 1.5 MTok × $2 = $3.00
• Output: 0.1 MTok × $10 = $1.00
• Total: about $4.00

Example 2: the same task on Opus 5.5.

• Input: 1.5 × $4 = $6.00
• Output: 0.1 × $20 = $2.00
• Total: about $8.00, double Sonnet for the same tokens.

Example 3: a month of light use. Three sessions like Example 1, four days a week, about 17 working days a month: 51 sessions × $4 = about $204 on pay-as-you-go. On a subscription, the same work is covered by the plan price as long as you stay within its limits.

Simple cost table

Prices as of Oct 21, 2026. Check the official page before you rely on them.

Is Claude Pro or Max better for Claude Code?

Pro suits part-time coding; Max suits people who hit Pro's limits. Max gives 5x or 20x Pro's usage per five-hour session, higher output limits, priority access at busy times, and monthly API credits ($100 on Max 5x, $200 on Max 20x) for building on the Claude Platform. If you rarely see a limit message on Pro, Max won't save you anything.

How do I check my Claude Code usage and cost?

Run /usage inside Claude Code. On a Console account it shows a Session block with token counts and an estimated "Total cost." On Pro and Max, the docs note that the session cost figure "isn't relevant for billing purposes" because usage is included; you see plan usage bars and a breakdown instead. For the official bill, use the Usage page in the Claude Console.

How can I lower my Claude Code bill?

From Anthropic's cost docs (external):

1. Use Sonnet for most coding. The docs say Sonnet "handles most coding tasks well and costs less than Opus." Switch with /model.
2. Run /clear between unrelated tasks. Stale context is re-sent with every message.
3. Keep CLAUDE.md short. It loads at every session start; the docs suggest under 200 lines and moving workflow detail into skills.
4. Turn off MCP servers you aren't using (/mcp).
5. Lower the effort level with /effort for simple jobs, since thinking tokens bill as output.
6. Write specific prompts. "Add input validation to the login function in auth.ts" costs less than "improve this codebase."
7. Set a spend limit. Use workspace limits in the Console, or a monthly limit on usage credits.

Does Claude Code cost money when idle?

A little. The docs say background jobs, such as summarizing past conversations for --resume, typically use "under $0.04 per session." Scheduled loops and agent teammates keep using tokens until you stop them.

Cheap to run, still worth checking

A coding agent can ship a working app in an afternoon, and it can also leave an API key in your frontend or a database table open. Whatever plan you're on, keep keys on the server, review what the agent changed, and read its permission settings before you approve everything by reflex. If your app calls Claude, see Claude API key security and what a .env file is.

FAQ

Is Claude Code free? No. It isn't included in the Free plan. The cheapest way to get it is Claude Pro at $20/month ($17/month billed annually), or pay-as-you-go through a Console account.

Is Claude Code included in Claude Pro? Yes. Anthropic lists Claude Code in Pro, Max, Team and Enterprise. It shares the same usage limits as your chats.

How much does Claude Code cost per month? On a plan, $20 (Pro) or from $100 (Max). On pay-as-you-go, it depends on tokens; Anthropic reports $150–250 per developer per month on average across enterprise deployments.

Is Claude Code cheaper on a subscription or the API? For steady daily use, a subscription is usually more predictable. For rare or bursty use, pay-as-you-go can cost less. Track a week with /usage and compare.

What happens when I hit my Claude Code limit? You wait for the five-hour or weekly window to reset, upgrade, or turn on usage credits billed at standard API rates.

Where are the per-model API prices? In our Claude API pricing guide, which covers every current model, batch and caching discounts and tool fees.

Vibe-coded your app? Scan it before launch.

Claude Code makes building fast. Before real users arrive, run a free Greenlight My App scan of your live site. It's automated, read-only and educational, not a security audit: it checks HTTPS, common security headers, exposed .env or .git files, public source maps and common secret key patterns in the scripts your homepage loads. "Nothing found" isn't a guarantee that nothing is exposed.