Free tool · Runs in your browser
Claude Code cost calculator
Claude Code has no price of its own. It comes with Claude Pro at $20 a month and with Max from $100 a month, or you pay per token through the API. Enter how you work and the calculator shows what the same usage costs on the API, next to both plan prices.
The rates are Anthropic's published prices in US dollars. Nothing you enter leaves your browser.
Last checked 24 September 2026
What Claude Code reads: your files, instructions and the conversation so far.
What it writes back: the answer and the edits.
Context that repeats from step to step, such as a CLAUDE.md or files it keeps rereading.
Per step
$0.60
Per day
$6.00
Per month on the API
$126.00
The plans
Claude Pro $20.00 a month · Claude Max from $100.00 a month
At this usage the API costs more than the Max entry price. A plan caps what you pay; whether your work fits its limits only shows in use.
Claude Sonnet 5: 200,000 + 20,000 tokens. Prices in US dollars, checked 24 September 2026: API rates, plan prices.
The monthly figure
How much does Claude Code cost per month?
On a plan, $20 with Pro or from $100 with Max, and that figure does not move however much you work. On the API it depends on how you work, so the table prices four typical ways of working at list price.
| How you work | Assumption | Haiku 4.5 | Sonnet 5 | Opus 5.5 |
|---|---|---|---|---|
| Occasional use | 5 steps a day, 100k in, 8k out | $7.14 | $14.28 | $26.88 |
| Daily work on one project | 20 steps a day, 200k in, 20k out | $65.52 | $131.04 | $248.64 |
| Heavy days, large codebase | 60 steps a day, 300k in, 30k out | $294.84 | $589.68 | $1,118.88 |
| An agent on a schedule | 200 steps a day, 150k in, 10k out | $386.40 | $772.80 | $1,444.80 |
For comparison, Anthropic's own documentation reports that across enterprise deployments the average is around $13 per developer per active day and $150 to $250 per developer per month.
Plans
Claude Code pricing: every plan and what it costs
Claude Code has no plan of its own. It is included from Pro upwards, and a team plan buys it per seat.
| Plan | Monthly billing | Annual billing | Claude Code |
|---|---|---|---|
| Free | $0 | $0 | Not included |
| Pro | $20 a month | $17 a month, $200 upfront | Included |
| Max | From $100 a month | Monthly only | Included |
| Team, Standard seat | $25 per seat | $20 per seat | Included |
| Team, Premium seat | $125 per seat | $100 per seat | Included |
| Enterprise | $20 per seat, billed annually, plus usage | Included |
Max is listed only as "from $100". The pricing page offers it at 5x or 20x Pro usage and does not print a second price, so this page does not either.
Free
Is Claude Code free?
No. The Free plan does not include it. The cheapest way in is Pro at $20 a month. New API accounts get a small amount of free credit to test with, which covers a first look at Claude Code on the API and not much more.
Per token
Claude Code API cost per token
On the API every model has an input price and a higher output price, both per million tokens. Reading from the prompt cache costs a fraction of the input price, and that fraction is not the same on every model.
| Model | Input | Output | Cache read | Batch in / out |
|---|---|---|---|---|
| Claude Fable 5.1 | $10.00 | $50.00 | $0.250 | $5.00 / $25.00 |
| Claude Opus 5.5 | $4.00 | $20.00 | $0.200 | $2.00 / $10.00 |
| Claude Opus 5 | $5.00 | $25.00 | $0.500 | $2.50 / $12.50 |
| Claude Sonnet 5 | $2.00 | $10.00 | $0.200 | $1.00 / $5.00 |
| Claude Haiku 4.5 | $1.00 | $5.00 | $0.100 | $0.500 / $2.50 |
Plan or API
When is the API more expensive than Pro or Max?
The API wins on light use and a plan wins on steady use. The table shows how many working steps a day it takes before the API bill passes each plan price, at the daily-work step size with 80% caching.
| Model | Cost per step | API passes Pro ($20) | API passes Max ($100) |
|---|---|---|---|
| Claude Haiku 4.5 | $0.16 | 7 steps a day | 31 steps a day |
| Claude Sonnet 5 | $0.31 | 4 steps a day | 16 steps a day |
| Claude Opus 5.5 | $0.59 | 2 steps a day | 9 steps a day |
What no table can show is whether that work fits inside a plan. Anthropic does not publish the Pro and Max limits in tokens, so treat the plan price as the most you pay, not as a promise that your work fits.
Caching
How much prompt caching saves
Claude Code sends your whole conversation with every request and reads the repeated part from the cache. The more of each step comes from the cache, the less a month costs. Daily-work step size:
| Share from cache | Sonnet 5 per month | Opus 5.5 per month |
|---|---|---|
| 0 % | $252.00 | $504.00 |
| 50 % | $176.40 | $344.40 |
| 80 % | $131.04 | $248.64 |
| 90 % | $115.92 | $216.72 |
Caching is automatic. What breaks it is a pause longer than the cache lifetime, after which the next message reprocesses the full context at the input price.
Teams
What Claude Code costs a team
On a team plan each member's usage draws on a seat allowance that resets on a rolling five-hour window and a weekly window. How large the allowance is depends on the seat tier, Standard or Premium.
| Seats | Standard, monthly | Standard, annual | Premium, monthly | Premium, annual |
|---|---|---|---|---|
| 3 | $75 a month | $60 a month | $375 a month | $300 a month |
| 5 | $125 a month | $100 a month | $625 a month | $500 a month |
| 10 | $250 a month | $200 a month | $1,250 a month | $1,000 a month |
Measure
How to check what you actually spend
In Claude Code, run /usage. On the API, its Session block shows the total cost of the current session, computed from token counts at list price, plus a line on how much of your input came from the cache. On Pro or Max the same screen shows your plan usage bars instead, because the session figure is not what you are billed.
For the authoritative number on the API, the Usage page in the Claude Console is the bill. Run a week of normal work, read the figure, and you have better input for the calculator than any assumption on this page.
Habits
Seven ways to spend less on Claude Code
Context
Clear between tasks
Run /clear when you switch to unrelated work. Old context is paid for again on every later message.
Model
Default to Sonnet
Sonnet handles most coding work at a fraction of Opus. Switch to Opus with /model for the hard decisions only.
Setup
Keep CLAUDE.md short
It loads into every session. Anthropic suggests under 200 lines, with specialised instructions moved into skills.
Prompts
Ask for something specific
A vague request makes Claude read the whole project. Name the file and the change and it reads far less.
Planning
Plan before you build
Plan mode lets Claude propose an approach before it writes code, which avoids paying twice for work in the wrong direction.
Thinking
Lower the effort on easy tasks
Thinking tokens bill as output. A lower effort level for simple edits cuts the most expensive part of a step.
Tools
Switch off unused MCP servers
Each configured server puts its tool names and instructions into the context. Disable the ones you are not using with /mcp.
All seven come from Anthropic's Claude Code cost documentation. Claude Code cost documentation
How it works
One working step is one request: Claude Code reads your files and instructions as input tokens and writes its answer and edits as output tokens. The calculator multiplies each by the model's rate per million tokens, then by your steps per day and working days per month.
The cached share of your input is billed at the model's cache read rate, which is a tenth of the input price or less. The one-off cost of writing the cache is left out, so the estimate errs slightly low, never high.
What it cannot tell you is whether your usage fits inside Pro or Max. Anthropic does not publish those limits in tokens, so the plan price is the most you pay on a plan, not a promise that your work fits.
Frequently asked
Is Claude Code free?
No. It is not part of the Free plan. The cheapest way in is Claude Pro at $20 a month, or $17 a month when you pay $200 for the year upfront.
How many tokens does one Claude Code step use?
It depends on how much context the step carries, not on how long your question is. A short question in a project with many open files reads far more input than a long question in an empty folder. The default of 200,000 input and 20,000 output tokens is an assumption for a mid-sized project, so replace it with numbers from your own usage.
Is the API or a subscription cheaper?
For regular daily work, almost always the subscription, because it caps what you pay. The API suits irregular work, or automation whose volume you control. On the API there is no cap, so an unexpected bill is possible there and nowhere else.
Why is the Batch API discount not included?
The Batch API halves input and output prices, but only for requests processed asynchronously in batches. An interactive Claude Code session is not a batch, so the discount does not apply to the work this calculator prices.
How much does Claude Code cost per month?
On a plan, $20 with Pro or from $100 with Max. On the API it depends on how much you work: Anthropic's documentation reports an average of around $13 per developer per active day and $150 to $250 per developer per month across enterprise deployments.
How much do Claude Code tokens cost?
Per million tokens on the API: Claude Sonnet 5 costs $2 input and $10 output, Claude Opus 5.5 $4 and $20, Claude Haiku 4.5 $1 and $5. Input read from the prompt cache costs a tenth of the input price or less.
How do I see what Claude Code has cost me?
Run /usage in Claude Code. On the API its Session block shows the cost of the current session at list price. On Pro or Max it shows your plan usage instead, because the session figure is not what you are billed.
Is Claude Code included in the Team plan?
Yes. A Standard seat costs $25 a month, or $20 billed annually, and a Premium seat $125, or $100 billed annually. Each member's usage draws on a seat allowance that resets on a rolling five-hour window and weekly.
More free tools
All free toolsThe tool is the small part of the bill
The larger part is the time until you build something with it that you want to keep. That is what The Claude Code System works on, four weeks of live sessions.
See the courseFor information only. The results are estimates and templates that may not fit your situation, and ENLIX accepts no liability for decisions based on them. Check anything important with Anthropic's current pricing page before you commit to a plan.