How Much Does Claude Code Cost? Plans, API Prices and How to Spend Less

Claude Code costs $20 a month on the Claude Pro plan ($17 a month if you pay yearly), $100 a month on Max 5x, or $200 a month on Max 20x. Teams pay $25 per standard seat monthly ($20 yearly). The free Claude plan does not include Claude Code. You can also skip a subscription and pay per token through a Claude Console (API) account. Anthropic says enterprise users average about $13 per developer per active day. To spend less, clear context between tasks, use Sonnet instead of Opus for everyday work, keep your CLAUDE.md short, and check /usage often. All prices here are as of October 2026, taken from Anthropic's own pages.

How much does Claude Code cost per month?

Claude Code is included in every paid Claude plan. Claude Code and Claude chat share the same usage limits.

Team plans are for 2 to 150 people. Prices exclude tax, and Anthropic notes that mobile app pricing may differ from web prices.

How do Claude Code usage limits work?

Every plan has usage limits that reset on a rolling five-hour session window. Paid plans add weekly limits on top. Max plans have a weekly limit across all models too.

When you hit a limit, you have three choices:

1. Wait for the window to reset. The message tells you when.
2. Turn on usage credits. This lets you keep working past your limit, billed at standard API rates. Run /usage-credits in Claude Code to manage them. On Pro and Max you can set a monthly spend limit.
3. Upgrade to a bigger plan.

Good to know: if you see "You've hit your Opus limit", switching to a model outside that family with /model keeps you working. A general session or weekly limit applies to all models, so switching does not help there.

How much does Claude Code cost with an API key?

With a Claude Console account, you pay for each token you use. There is no monthly fee. These are Anthropic's list prices per million tokens (MTok) for the current models:

Claude Code sends a lot of repeated context, such as your files and conversation history. Prompt caching makes those repeats much cheaper, and Claude Code uses it automatically.

What does that mean in real money? Anthropic's own figures from enterprise deployments:

• About $13 per developer per active day
• About $150 to $250 per developer per month
• Under $30 per active day for 90% of users

Your number depends on your model, project size and habits.

Warning: if the ANTHROPIC_API_KEY environment variable is set, Claude Code can use that key instead of your subscription. Anthropic's support page says this means API charges instead of plan usage. If you pay for Pro or Max, check that no old key is set.

Is Claude Pro or Max better value for Claude Code?

It depends on how often you hit limits.

A simple test: start on Pro. If you hit the limit most days, Anthropic's support article suggests Max 5x as the next step. Compare the upgrade to what usage credits would cost you at API rates.

If you are still choosing between tools, our Claude Code vs Cursor vs GitHub Copilot comparison lists all three vendors' prices side by side.

How do I check how much Claude Code I am using?

Run this inside a session:

/usage

/cost and /stats are aliases for it. On a Pro, Max, Team or Enterprise plan it shows:

• Plan usage bars and activity stats
• Attribution: what share of recent usage came from skills, subagents, plugins and each MCP server
• Recommendations: specific fixes, such as "Trim CLAUDE.md" or "Compact before you step away", each with the share of usage it addresses

Press d or w to switch between the last 24 hours and the last 7 days. If the footer shows a hint like f to fix 2 of 3, press f and Claude Code applies the fixes it can (this needs v2.1.305 or later).

API users also see a session cost estimate at list price. For exact billing, check the Usage page in the Claude Console.

How can I make Claude Code cheaper?

Token costs grow with context size. The more Claude has to read, the more you pay or the faster you hit limits. These tips come from Anthropic's cost guide.

1. Clear between tasks. Run /clear when you switch to unrelated work. Old context is re-sent with every message. Use /rename first if you want to find the session later with /resume. Clearing costs nothing.

2. Pick the right model. Anthropic says Sonnet handles most coding tasks well and costs less than Opus. Save Opus for complex architecture or multi-step reasoning. Switch with /model. For simple subagent tasks, set model: haiku in the subagent's settings.

3. Lower the effort level. Thinking tokens are billed as output tokens. For simple tasks, run /effort and choose a lower level.

4. Keep CLAUDE.md short. It loads into every session. Anthropic suggests under 200 lines. Move long procedures into skills, which load only when used. Our CLAUDE.md guide shows how.

5. Compact with focus. /compact Focus on code samples and API usage tells Claude what to keep. Note that compacting a huge conversation is itself a large request, so /clear is cheaper when you do not need continuity.

6. Trim MCP servers. Run /mcp and disable servers you are not using. Anthropic notes that CLI tools like gh or aws are more context-efficient than MCP servers.

7. Write specific prompts. "Improve this codebase" makes Claude scan widely. "Add input validation to the login function in auth.ts" keeps reads small.

8. Plan first. Press Shift+Tab to switch to plan mode for complex work. Agreeing on an approach first avoids paying for rework. Press Esc early if Claude heads the wrong way, and use /rewind to go back.

9. Send noisy work to subagents. Running tests or reading logs can flood your context. A subagent keeps that output in its own context and returns a summary. Its requests still count, so give it a smaller model.

10. Filter with hooks. A hook can cut a 10,000-line log down to the error lines before Claude reads it. See our guide to Claude Code skills, hooks and slash commands for examples.

11. Watch long sessions. A session open all day re-sends its whole history with each request. Your first message after a long break can miss the cache and reprocess everything. Scheduled /loop tasks also fire while you are away.

12. Be careful with agent teams. Anthropic says agent teams use about 7 times more tokens than standard sessions when teammates run in plan mode. Keep teams small and shut teammates down when done.

What else costs tokens in the background?

Some background work uses a small amount of tokens even when you are idle, such as summarizing past conversations for claude --resume. Anthropic says this is typically under $0.04 per session. Prompt suggestions also send a short request after each reply. You can turn them off.

Can I offload work to other AI tools to save money?

Yes. Not every task needs Claude Code's full agent. Routine jobs like dependency bumps or simple bug fixes can go to other coding agents, so your Claude limits last longer for the hard work. We walk through one such setup in how to save money with Google Jules and Claude.

How do teams control Claude Code spending?

• Team and Enterprise plans: each member draws from a per-seat allowance that resets on a five-hour and a weekly window. Admins can turn on usage credits and set spend limits per organization, group or member in the claude.ai admin console.
• Claude Console (API): set workspace spend limits. A workspace called "Claude Code" is created on first login for cost tracking.
• Cloud providers (Bedrock, Google Cloud, Microsoft Foundry): spend controls live in your cloud billing console.

Frequently asked questions

Is Claude Code free? No. The free Claude plan does not include Claude Code. The cheapest option is Pro at $20/month, or $17/month billed yearly, as of October 2026. A Console account has no monthly fee but charges per token.

What is the difference between Max 5x and Max 20x? Max 5x costs $100/month and gives 5 times Pro's per-session usage. Max 20x costs $200/month and gives 20 times Pro's per-session usage. Both include Claude Code.

Do Claude chat and Claude Code share the same limits? Yes. All activity in both counts against the same limits on Pro and Max.

What happens when I hit my Claude Code limit? Wait for the five-hour window to reset, turn on usage credits billed at API rates, or upgrade your plan. Run /usage-credits to manage credits.

How do I see my Claude Code costs? Run /usage (or /cost). Subscribers see plan usage bars and recommendations. API users see a session cost estimate, with exact billing in the Claude Console.

Which model is cheapest for Claude Code? On the API, Haiku 5.5 is the cheapest current model. For most coding, Anthropic recommends Sonnet as the balance of cost and capability, with Opus for complex reasoning.

Learn to use Claude Code efficiently, for free

The cheapest token is the one you never send. Our free course Claude Code from Zero teaches the habits that keep costs down: clean context, clear prompts and a lean setup. New to the tool? Start with our beginner's guide to installing Claude Code.

Sources

• Claude pricing
• Claude Max plan page
• Claude Help Center: What is the Max plan?
• Claude Help Center: Using Claude Code with your Pro or Max plan
• Claude Code docs: Manage costs effectively
• Claude Code docs: Commands
• Claude Code docs: Advanced setup (authentication)
• Claude API docs: Pricing