Guide · Claude Code cost
How much does Claude Code cost per developer?
About $13 per active day and $150 to $250 a month, by Anthropic's own figures across enterprise deployments. Nine in ten developers stay below $30 on a day they use it. Those figures are per-token costs at list price. Your own team's spread depends on the model, the codebase and how many sessions run at once, so measure it per person.
What does Anthropic say a developer costs?
Anthropic publishes three figures for Claude Code billed per token:
| Measure | Anthropic's figure |
|---|---|
| Average, per developer per active day | About $13 |
| Per developer per month | $150 to $250 |
| 90% of users, per active day | Below $30 |
Anthropic adds that costs per developer vary widely with the model, the size of the codebase, and habits such as running several instances or automation. It suggests starting with a small pilot group to set a baseline before a wider rollout.
Why do two developers cost so differently?
Token cost follows the context a developer carries, not the hours they spend.
Anthropic's own documentation names the causes:
- The model. Sonnet costs less than Opus per token. On an API key, Anthropic traces unexpectedly high spend usually to long sessions that were never cleared or to Opus left as the default model.
- Long sessions. Claude Code sends the whole conversation with every request. A one-line question in a session open all day still pays to re-read that day.
- Cache misses. The first message after a break longer than the cache lifetime processes the whole context again. On an API key that lifetime is five minutes by default.
- Work running in parallel. Every subagent sends its own requests. Agent teams use about seven times the tokens of a standard session when teammates run in plan mode, and a scheduled task fires even while its session sits idle.
What each kind of token costs, model by model, is in the Max plan vs API guide.
Does a subscription seat cost the same?
No. A Claude seat is a flat monthly fee, and use inside its allowance is not billed per token. A seat stops when it fills its 5-hour or weekly window, rather than costing more.
Anthropic's figures describe traffic billed per token, so they price a seat-holder's work, not their seat. A seat pays off for a developer whose traffic would cost more than its fee at list price. The break-even, plan by plan, is in is a Claude Max plan cheaper than the API.
How do you measure cost per developer on your own team?
Anthropic's tools depend on how your team signs in to Claude Code:
| How you pay | Where cost per person shows | What it leaves out |
|---|---|---|
| Claude Team or Enterprise seats | The spend report in org analytics, per user and model, as CSV | Use inside the seat allowance, which is not metered in dollars |
| Claude Console API | The Console's Claude Code dashboard and the Claude Code Analytics API, per member | Developers who signed in another way |
| Bedrock, Vertex or Foundry | OpenTelemetry from each laptop, or a gateway | Anthropic's dashboards, which do not see this traffic |
On one machine, /usage prices the current session at list price. It is an
estimate, and it covers that laptop only. Every one of these reads Claude alone, so a
team on Codex as well has a second bill to join by hand.
How does Selan show what each developer costs?
Selan connects the seats and API keys your company already pays for. Every Claude Code, Codex and Claude Desktop request is recorded with who ran it, which account served it, which model answered and what it cost.
Usage, in the control app, ranks people by cost over the last 24 hours, 7 days or 30 days, in UTC. Click a name and the whole page narrows to that person. A seat's traffic is priced at what the same requests would have cost per token, next to what your keys actually cost. A request that cannot be priced shows as unknown, never as zero.
What the page shows, chart by chart, is in how to see Claude Code usage per developer. How seats and keys are connected is in Providers.
How do you cap what one developer can spend?
In Selan, set a spend limit per person, by day, week or month. A person whose field is blank inherits the company limit, so one default covers the team and an exception covers whoever needs one.
The limit follows the person across every pay-per-use key you added. It binds the moment it is reached, and resets at the end of its calendar day, week or month in UTC. Requests use the seats first and reach a key only when the seats are capped.
Each developer sees their own spend against their own limit on the statusline, described in the statusline docs. Limits are on the spend page, and setup starts at the docs.
Free for up to three seats, and the Free plan needs no card.
The short version
- Anthropic puts Claude Code at about $13 per developer per active day and $150 to $250 a month, billed per token, with nine in ten below $30 a day.
- The model, long sessions, cache misses and parallel agents decide where a developer lands.
- A seat is a flat fee that stops at its window; per-token figures price the work, not the seat.
- With Selan, every request carries a person's cost, and a limit per person caps what they spend on keys.
Anthropic's figures, the causes of high spend and its reporting tools are as Claude Code's cost documentation (code.claude.com/docs/en/costs) stated them on 17 October 2026. They change: check the date before quoting a figure. What Selan's Usage page and limits do is as this site's spend page and usage guide describe them.