Claude Code Pricing
> AI content disclosure: AI assisted with the comparison format; plan prices, usage rules, and API rates were rechecked on the linked official Claude pages on July 17, 2026.
Further reading: See the full Claude pricing guide
Claude Code pricing depends on how you authenticate and how heavily you work. Interactive Claude Code is included with paid Claude plans and shares their usage pool. API-key usage is pay-as-you-go at model token rates. These are separate billing paths, so the cheapest choice depends on session pattern, model selection, and whether predictable subscription spend matters more than exact usage billing.
Claude Code Pricing Models: Subscription vs API
| Access path | How you pay | Best fit | Main constraint |
|---|---|---|---|
| Claude Pro | $20/month or $200/year | regular interactive work | rolling session and weekly limits |
| Claude Max 5x | $100/month | frequent long sessions | 5x Pro capacity per session, still limited |
| Claude Max 20x | $200/month | daily heavy use | 20x Pro capacity per session, still limited |
| Claude API | per model token and paid tool | measurable automation or variable use | spend varies with context and output |
Prices above are US web prices rechecked July 17, 2026; mobile, tax, currency, and plan terms can differ. Claude Code, Claude web, desktop, and mobile draw from the same subscription usage pool. API billing does not consume that included pool.
Pro Plan for Claude Code: What You Get
Pro includes Claude Code and the broader paid Claude feature set with more usage than Free. It is the sensible baseline for a solo developer who codes interactively but can pause when a usage window resets. The plan does not represent unlimited tokens or a fixed number of prompts.
Usage depends on prompt length, conversation history, model, attachments, tool output, and task complexity. A compact bug fix may cost far less capacity than a repository-wide analysis. Check Settings > Usage instead of relying on an unofficial “messages per day” estimate.
Max Plan Token Limits vs API Pay-Per-Use
Max is offered as 5x and 20x Pro capacity per rolling five-hour session. Paid plans also have weekly limits. Max buys a larger included pool and higher output limits; it does not publish a permanent token allowance that applies equally to every workload.
API usage bills actual tokens. For estimates rechecked August 11, 2026, Sonnet 5 is $2/MTok input and $10/MTok output. Anthropic made these rates permanent on August 10; the previously announced $3/$15 September rate no longer applies. Opus 4.8 is $5/$25 and Haiku 4.5 is $1/$5. The official pricing table is the source of truth for model promotions, cache rates, and geography modifiers.
Cost Calculator: How Much Will You Actually Spend
For API use:
~~~text monthly cost = input MTok × input rate + output MTok × output rate + cache writes and reads + paid server tools + applicable geography or speed premiums ~~~
Example: 12 million Sonnet 5 input tokens and 2 million output tokens at the current Sonnet 5 rates cost $44 before caching or tools: 12 × $2 plus 2 × $10. This is an illustration, not a promise about how many coding tasks those tokens cover.
For subscriptions, track interruptions as well as dollars. If Pro repeatedly blocks work and the delay costs more than the upgrade, Max may be rational. If heavy use happens only a few days each month, Pro plus carefully capped extra usage or direct API billing may be cheaper.
Claude Code Free Trial & Free Credits
Claude's Free plan is not the same as a guaranteed Claude Code trial allowance. The current pricing page states that Claude Code is included in paid plans. Promotions and credits can appear, but they are time-bound and account-specific. Verify an offer inside the official billing interface and do not enter payment details through a third-party “student” or “free credit” page.
Claude Code Pro vs Max Plan Decision Guide
Choose Pro when interactive coding is regular but stoppable. Choose Max 5x when five-hour-session capacity is the recurring bottleneck. Consider Max 20x only when Claude is used across most of the workday and the larger included pool is demonstrably cheaper than delays or pay-as-you-go. Choose API billing for programmatic runs, auditable per-token cost, or workloads that do not fit interactive subscription limits. Set a spend cap before automation.
In Verdent
Verdent is a separate product and billing system, not a reseller of Claude subscriptions. Verdent offers built-in models from several providers and uses credits across its Desktop and VS Code products; its model-pricing page discloses provider-side token references. Compare workflows as well as price: Claude Code offers Anthropic's terminal agent, while Verdent emphasizes multiple isolated workspaces and parallel coding tasks.
Sources: Claude plans and pricing, Choose a Claude plan, Claude API pricing, Verdent model pricing.
Frequently Asked Questions
Is Claude Code included with Claude Pro?
Yes. Claude Code is included with paid Claude plans and shares the same usage pool as Claude on other supported surfaces.
Does Claude Max provide unlimited Claude Code use?
No. Max increases capacity to 5x or 20x Pro per session and also has weekly and other fair-use limits.
Can I pay for Claude Code through the API?
Yes. API-key usage is billed by model tokens and paid tools, separately from the included usage in a Claude subscription.
Is there a fixed Claude Code cost per task?
No. Cost or capacity depends on model, context, output, tool results, caching, and task complexity. Measure representative sessions.
Compare Claude Code costs with Verdent
Review Verdent’s model pricing to see how credits apply across Desktop and VS Code, then compare Claude Max usage if you are weighing subscription limits against coding workflows.