Key findings
- Claude plans (Pro, Max, Team, and Enterprise) include Claude Code. The Claude API bills per token through the Claude Console or a cloud provider.
- On paid plans, Claude Code shares one usage pool with Claude chat, so coding adoption draws on the same allowance as everyday use.
- Data terms follow the route. Anthropic does not train on Team, Enterprise, or API usage by default, and trains on Free, Pro, and Max data when the user's setting is on.
Choose how your company buys Anthropic before teams buy seats, because the purchasing route sets usage limits and data terms. Claude (the assistant), Claude Code (the coding agent), and the Claude API (the developer platform) share models and reach buyers through two routes.
Two purchasing routes
Claude Pro costs US$20 per month, or US$17 a month billed annually, and Max starts at US$100 per month. Team, for 2 to 150 seats, costs US$25 per Standard seat per month, or US$20 billed annually, and US$125 per Premium seat, or US$100 billed annually. Enterprise costs US$20 per seat per month billed annually plus all usage at API rates, with a 20-seat minimum self-serve and 50 sales-assisted. Claude Code is included in every paid plan.
The Claude API bills per million tokens. Opus 5.5, Claude Code's default model on paid plans and the API, costs US$4 input and US$20 output per million tokens. Claude Code can also run through Amazon Bedrock, Google Cloud, or Microsoft Foundry, and the cloud provider bills that usage.
Why it belongs at company level
A team can prefer Anthropic's models for API work and still choose another vendor's product for daily coding, or the reverse. The purchasing route, data terms, and administrator controls apply across Claude, Claude Code, and the API, so GILD covers them on the company page and each product page evaluates its own product.
Methodology
Prices and terms are Anthropic's own, from its pricing pages and documentation, checked on the dates shown.
Products in this piece
Sources
- Official documentation
- 6