Claude Code has no subscription of its own: it is included in Claude Pro ($20 a month), Max ($100 or $200), every Team seat and Enterprise, and it shares each plan's five-hour and weekly limits with chat. Without a plan it bills per token through a Claude Console account, where Anthropic reports an average of about $13 per developer per active day. Pick a plan by how often you hit limits, and an API key when you need hard cost control or scripted runs.
Claude Code pricing is the price of the Claude plan you sign in with: $20 a month on Pro, $100 or $200 a month on Max, from $20 a seat on Team, or pay per token through a Claude Console account. There is no separate Claude Code subscription, and the free Claude plan does not include it.
Checked October 2026, in US dollars before tax, from Anthropic's plan page, its help centre articles on Pro, Max, Team and Claude Code on Pro and Max, the Claude Code cost guide and the API pricing docs. The plan page shows local currency outside the US, so check it from your own location before you pay.
- Free plan: Claude Code not included
- Pro: $20 a month, or $17 a month billed annually. Shares limits with chat
- Max 5x: $100 a month for 5x the Pro per-session allowance
- Max 20x: $200 a month for 20x the Pro per-session allowance
- Team: $25 a seat ($20 annually) Standard, $125 ($100 annually) Premium
- Enterprise: $20 a seat a month billed annually, plus usage at API rates
- Console account: per token. Opus 5.5 is $4 in and $20 out per million tokens
This page covers only Claude Code: which plans include it, how its limits behave in the terminal, what an API key costs, and where the break-even sits. For every Claude plan, including the chat features, see our Claude pricing guide. If you are new to the tool itself, start with the Claude Code complete guide.
Which Claude plans include Claude Code?
Every paid plan includes it, and the Console counts as a plan of its own. The plan page marks Claude Code "No" on Free and "Yes" on Pro, Max 5x, Max 20x, Team and Enterprise, and Anthropic's Team and Enterprise article says it is included with every Team seat.
| Plan | US price (checked October 2026) | Claude Code | How much you get |
|---|---|---|---|
| Free | $0 | Not included | None |
| Pro | $20 a month, or $17 a month billed annually ($200 up front) | Included | Five-hour session limit plus a weekly limit, shared with chat |
| Max 5x | $100 a month (monthly billing only) | Included | 5x the Pro per-session allowance, plus a weekly limit |
| Max 20x | $200 a month (monthly billing only) | Included | 20x the Pro per-session allowance, plus a weekly limit |
| Team Standard seat | $25 a month, or $20 billed annually | Included | 1.25x Pro per session, per member |
| Team Premium seat | $125 a month, or $100 billed annually | Included | 6.25x Pro per session, per member |
| Enterprise | $20 a seat a month, billed annually, 20-seat minimum, plus usage at API rates | Included | Set by admins with spend limits |
| Claude Console (API) | Pay per token | Included | Your prepaid balance, workspace spend limit and rate limits |
Source: Anthropic help centre and claude.com/pricing, US dollars before tax, checked October 2026
Two billing details matter. Max is sold monthly only, while Pro and Team can be paid yearly. And Anthropic's Max article says the prices are for web subscriptions; app store prices can differ, so subscribing on the web is the way to get the listed price.
How Claude Code usage limits work
Limits are time windows, not a message count. Each plan has a session limit that resets every five hours and a weekly limit that resets at a fixed time assigned to your account, per the Pro article. Anthropic does not publish a number of prompts per window, because a request's cost depends on context size, model and task.
The rule that surprises people: Claude Code and Claude chat draw from the same pool. Anthropic's Claude Code on Pro and Max article says all activity in both tools counts against the same limits, and IDE use counts too. A long research chat in the morning leaves less for the terminal in the afternoon.
- Check where you stand. Run
/usagefor plan usage bars and a breakdown of what used them (skills, subagents, MCP servers), or/statusfor your remaining allocation. - Mid-task cutoffs. If you hit the five-hour limit mid-response, Claude Code may finish the current step and show "Usage limit reached · wrapping up". New messages after that wait for the reset.
- Model-specific limits. A session or weekly limit covers all models, so switching with
/modeldoes not help. After an "Opus limit" or "Sonnet limit" message, switching to another model family does, per the cost guide. - Fable models. On Pro and Team Standard seats, Fable 5 and 5.1 run on usage credits from the first message. On Max and Team Premium seats you can spend up to 50% of your weekly limit on Fable at no extra cost (help centre).
- 01Read the message
A session or weekly limit covers every model. An Opus or Sonnet limit only covers that family.
- 02Wait for the reset
The message shows the reset time. Recent versions can resume the interrupted task automatically after it.
- 03Turn on usage credits
Prepaid, billed at standard API rates, with a monthly spend cap. Run /usage-credits to open the settings.
- 04Move up a plan
Pro to Max 5x, or Max 5x to Max 20x, if you hit limits most weeks.
Usage credits keep you working past a limit and apply to both chat and Claude Code. They are prepaid, auto-reload is optional, and you can cap monthly spend. One cost detail from the cost guide: the prompt cache lasts an hour on a subscription but drops to five minutes once you draw on usage credits, so a break mid-session costs more on credits than inside the plan.
What Claude Code costs on an API key
On a Console account Claude Code bills every token at the API price list. Claude Code defaults to Opus 5.5 on every plan and on the API, per its model configuration docs, so check /model before a long session. Current rates per million tokens, checked October 2026: Opus 5.5 $4 input and $20 output, Sonnet 5.5 $2 and $10, Haiku 5.5 $0.10 and $0.50 (up to a 100,000-token prompt), Fable 5.1 $10 and $50. Cache reads, which make up most of a long session, cost $0.20 on Opus 5.5 and $0.10 on Sonnet 5.5.
Anthropic's cost guide gives the only published benchmark: across enterprise deployments, about $13 per developer per active day and $150 to $250 per developer per month, with 90% of users below $30 per active day. Those are full-time professional developers, so treat them as a ceiling for a solo builder rather than a target.
On a Console account, Claude Code creates a workspace called "Claude Code" the first time you sign in. Set a workspace spend limit there, leave auto-reload off while you learn your usage, and read the Console usage page for the real bill: the dollar figure in /usage is a local estimate at list price. Setup for the key itself is in our guide to getting a Claude API key.
Claude Code cost examples: subscription vs API
The break-even is simple to state: an API key wins when your monthly token bill would stay under the plan price, and a plan wins once it would pass it. The examples below put numbers on that. The token counts are illustrative inputs, not measurements; only the per-token rates are Anthropic's.
Illustrative task: a focused change in a mid-sized repo that takes about 20 requests, each re-reading roughly 50,000 tokens of cached context (1,000,000 cache-read tokens in all), adds 100,000 tokens of new context (billed as five-minute cache writes) and produces 40,000 output tokens, thinking included.
| Model | Cache reads (1M tokens) | New context (100K tokens) | Output (40K tokens) | Cost per task |
|---|---|---|---|---|
| Claude Opus 5.5 (default) | $0.20 | $0.50 | $0.80 | $1.50 |
| Claude Sonnet 5.5 | $0.10 | $0.25 | $0.40 | $0.75 |
| Claude Haiku 5.5 | $0.01 | $0.0125 | $0.02 | about $0.04 |
| Claude Fable 5.1 | $0.25 | $1.25 | $2.00 | $3.50 |
Scale that to a month and the plan choice falls out. The last column compares each total with the plan prices above.
| Illustrative usage | Tasks a month | On Opus 5.5 | On Sonnet 5.5 | Cheapest fit |
|---|---|---|---|---|
| Light: 2 tasks a day, 8 days a month | 16 | $24 | $12 | Pro, or a capped Console balance |
| Medium: 6 tasks a day, 20 days a month | 120 | $180 | $90 | Max 5x |
| Heavy: 15 tasks a day, 22 days a month | 330 | $495 | $247.50 | Max 20x |
The pattern holds across models: light, irregular use is cheaper on a capped Console balance or Pro; daily use on Opus passes the Max 5x price quickly; and switching routine work to Sonnet with /model halves the API bill. If you would sooner try before paying anything, our Claude Code free options guide covers trial credits and programs.
How to estimate your own Claude Code cost
One week of real usage beats any table. This takes about five working days and needs no tools beyond Claude Code.
- 1Start on Pro or a small Console balance
Pro if you also use Claude chat; a Console balance with auto-reload off if you only want the agent.
- 2Work normally for five days
Same repos, same kind of tasks. Do not change habits for the test.
- 3Read /usage at the end of each day
On a plan, note how much of the session and weekly bars you used. On the API, note the session cost.
- 4Note every limit you hit
A limit on day two of a weekly window means Pro is too small for your pattern.
- 5Multiply to a month
API: daily cost times working days. Plan: how many weeks you would hit the weekly limit.
- 6Pick the plan just above your number
Max 5x if Pro runs out weekly; Max 20x if Max 5x does; stay on the API if the total is under $20.
Claude Code pricing for teams and Enterprise
For a company, the choice is seats or tokens. On Team and Enterprise seats, each member's Claude Code usage draws from a per-seat allowance on a five-hour and a weekly window, shared with Claude chat, and the allowance is not pooled across the team. Admins manage it in the claude.ai admin console: turn on usage credits so members can work past their allowance, and set spend limits for the organization, a group or one member.
- Team: 2 to 150 members, Standard and Premium seats mixed freely. Give Premium seats to the people who live in Claude Code. Team's monthly API credits ($20 per Standard seat, $100 per Premium, pooled up to $500) cannot be spent in interactive Claude Code (help centre).
- Enterprise: $20 a seat a month billed annually with a 20-seat minimum, plus usage billed at API rates, per the plan page. Usage-based Enterprise plans have no per-seat usage limits.
- Console for a team: per-token billing for everyone, workspace spend limits, and a per-user dashboard. The cost guide suggests 200,000 to 300,000 tokens per minute per user for teams of 1 to 5, falling to 10,000 to 15,000 for 500 or more, because fewer people run it at once in big teams.
- Wants a predictable monthly bill per person
- Also uses Claude chat every day
- Can give heavy users Premium seats
- Prefers limits over invoices that move
- Runs Claude Code in CI or scripts
- Has uneven usage across members
- Needs hard caps per workspace
- Wants cost per user from an API
If you plan to ship a product rather than only use the tool, AI SaaS Builder, one of our programs, has lessons on the Claude Code CLI and on prompt caching inside a full build. For how Claude Code compares with an editor that bills differently, see Claude Code vs Cursor, and to set it up, install Claude Code step by step.
Claude Code pricing: FAQ
How much does Claude Code cost?
Claude Code has no price of its own. It comes with a Claude plan: Pro at $20 a month ($17 a month billed annually), Max at $100 or $200 a month, a Team seat from $20, or Enterprise at $20 a seat plus usage. Without a plan, a Claude Console account bills it per token at API rates. Prices are in US dollars before tax, checked October 2026.
Is there a separate Claude Code subscription?
No. Anthropic sells no Claude Code-only plan. Claude Code is included in Pro, Max, every Team seat and Enterprise, and it shares the plan's usage limits with your chats in Claude. The free Claude plan does not include it. If you do not want a subscription, sign in with a Claude Console account and pay per token instead.
What are the Claude Code usage limits on the Pro plan?
Pro has a session limit that resets every five hours and a weekly limit across all models, and Claude Code shares both with Claude chat. Anthropic does not publish a fixed number of prompts, because usage depends on context size, model and task. Run /usage or /status inside Claude Code to see where you stand. When you hit a limit you can wait, turn on usage credits or move to Max.
Is the Claude Max plan worth it for Claude Code?
It is worth it when you hit Pro limits most weeks. Max 5x costs $100 a month for five times the Pro per-session allowance, and Max 20x costs $200 for twenty times. Max also lets you spend up to half your weekly limit on Fable models without usage credits. If you only code a few hours a week, Pro or pay-per-token usage costs less. Checked October 2026.
How much does Claude Code cost with an API key?
With a Console account or API key you pay per token. Opus 5.5, Claude Code's default model, costs $4 per million input tokens and $20 per million output tokens; Sonnet 5.5 costs $2 and $10. Anthropic's cost guide reports an average of about $13 per developer per active day across enterprise deployments, with 90% of users under $30. Your figure depends on model and session length.
Can I use the monthly API credits from Max or Team in Claude Code?
Not in interactive Claude Code. Anthropic's help centre says the monthly API credits on Max and Team cannot be used in Claude Code in the terminal, IDE, desktop or web, or for extra usage. They do apply to claude -p runs made with an API key from the Console organization you linked, and to the API, Managed Agents and the Agent SDK.
Does a Team Standard seat include Claude Code?
Yes. Anthropic's Team plan article says Claude Code is included with every Team seat. A Standard seat costs $25 a month, or $20 billed annually, with 1.25 times the Pro per-session allowance. A Premium seat costs $125, or $100 billed annually, with 6.25 times. Admins can mix both seat types in one team of 2 to 150 members.