Claude Code Pricing: Subscriptions, API Costs, and Limits

Claude Code pricing depends on how you plan to use it. You can access it through a Claude Max subscription, which bundles in a usage allowance, or by connecting it to the Anthropic API and paying for what you use. The right choice depends on how much coding work you send through it and whether you need steady access or occasional bursts.

Whichever plan you pick, the work itself — your project context, your coding rules and where each task stands — lives outside the billing plan. Tools like fenbs keep that record in one place, so if a budget change leads you to a different subscription tier or a different assistant, your context travels with you.

What Claude Code is and what triggers a charge

Claude Code is a command-line coding assistant that runs in your terminal, reads your local files and sends those files to the Claude model for analysis. It is not a website tab: it opens a session in your project folder and uses the model to read, write and edit code on your behalf.

Every exchange sends tokens to the API: the files you share, your instructions and the model's reply. Token consumption is the base unit for any Claude billing. Subscription plans convert that into an included usage allowance rather than a per-token charge, but the same underlying token use drives the cost on either path.

For the current plan names, allowances and included features, check the Anthropic pricing page↗ (opens in a new tab) directly. Plan details change, and a summary from a roundup can be months out of date.

Subscription versus API billing: which fits your work

A subscription gives you predictable monthly cost. You pay a fixed amount and use Claude Code within the included allowance. If your usage is fairly steady and you are doing your own work rather than building a product for others, a subscription is the simpler option to manage.

API billing charges per token. You connect Claude Code to your API key, and each session generates a bill proportional to the code and context you send. This option works well when usage is irregular, when you need a higher rate limit than a subscription includes, or when you are automating Claude Code inside a workflow that another tool controls.

Claude Max is the subscription tier aimed at heavy Claude Code users. It includes a larger usage budget than Claude Pro and carries a higher price. Check the Claude plans page↗ (opens in a new tab) for the current allowances, what a limit hit looks like in practice, and whether the plan throttles or stops access when the budget runs out.

Is there a free trial for Claude Code

Claude Code is not part of the free Claude tier. You need either a paid subscription that includes it or an Anthropic API account with a funded credit balance. Anthropic occasionally offers trial credits for new API accounts; check the Anthropic website for any current offer, because trial terms change.

If you want to understand what Claude Code does before committing, what is Claude Code explains the workflows it handles well, and how to install Claude Code covers setup so you know what you need in place before the first session.

Usage limits and what happens when you reach them

Claude Code sessions can consume tokens quickly when they involve large files, long conversations or repeated edits across a codebase. If you hit a usage limit mid-session, the most common outcome is a pause until the window resets, not an error that loses your work.

To work within limits, share only the files the task needs rather than opening your whole repository. You can use claude-code-sandbox mode to experiment without committing, and worktrees to keep parallel tasks in separate branches so each session is focused. Smaller, focused sessions stay within a budget more reliably than long exploratory ones.

For the current limit numbers, rate resets and what each plan does when you exceed the allowance, check Anthropic's help center directly. Plans have changed more than once, and a specific number quoted elsewhere may no longer be current.

API billing for teams and products

If you are building a product that calls Claude Code, or if several engineers in your company use it through a shared API key, API billing is the standard path. You can set spending limits in the Anthropic API console to prevent an unexpected month-end bill.

Anthropic offers enterprise agreements for organizations with high or predictable volume. Those conversations happen directly with Anthropic. For the current enterprise contact point, see the Claude docs.

How to do this with fenbs

1. Create a free account and add your project. A personal board is free, no credit card required.
2. Add your current tasks, Decisions and rules (style guide, the libraries you use, test requirements), and a Where we left off note with where each task stands.
3. Connect Claude Code over MCP from Integrations. Follow the guide there for current setup steps.
4. Ask Claude Code to read the fenbs project context at the start of each session. Before the session ends, ask it to save its progress to the same project.
5. If a plan change leads you to switch assistants, connect the new one to the same fenbs project. Your context and task state are already there.

For choosing between models in a session, see the claude-code-router guide on the Anthropic docs site. For a comparison of Claude Code's debugging strengths against other tools, see claude for debugging code.

When to go to the Anthropic page directly

Plan names, included allowances, trial offers and enterprise tiers change regularly. Before making a plan decision based on a specific price or limit, open anthropic.com/pricing↗ (opens in a new tab) and verify the current state. This page explains how the billing model works so you know what questions to ask; Anthropic's page carries the authoritative numbers.