Claude Max and Team plans now include monthly API credits, and interactive Claude Code can't spend them

Anthropic's release notes for October 7, 2026 carry two entries. One is Claude Haiku 5.5. The other is a billing change that matters more to anyone who pays for a Claude subscription and also writes code against the API: Max and Team plans "now include monthly API credits for running your own apps and agents on the Claude Platform." The help article explaining the program is specific about amounts, scope and expiry, and the scope is narrower than the headline.

What each plan gets

The amounts, per Anthropic's help article:

Team credits are pooled into one balance, capped at $500 a month and recalculated from the seat count at each deposit. Anthropic's own example: three Standard and two Premium seats yield $260 a month, and adding one Premium seat raises that to $360 the next cycle. Discounted Nonprofit and Scientist Team plans get the same per-seat amounts. The article is blunt about everyone else: "Free, Pro, and Enterprise plans aren't eligible."

You have to be on an eligible plan for seven days first, and you claim the credit on claude.ai in a web browser by linking a Claude Console organization under Settings > Billing (Max) or Organization settings > Billing (Team). No payment method on the Claude Platform is required.

What the credits cover, and what they don't

The credits work on any available Claude model through the Messages and Message Batches APIs, the Console Playground, Claude Managed Agents and the Claude Agent SDK. They also cover claude -p when it runs with an API key from the linked organization.

They do not cover interactive Claude Code in the terminal, IDE, desktop or web; runs started by the Claude Code GitHub Action, an IDE extension or the desktop app; extra usage in Claude, Claude Code or Claude Cowork; or Claude on Amazon Bedrock, Google Cloud Vertex AI or Microsoft Foundry. The credit is also not a raise in your plan limits: the article says credits don't change usage limits in Claude, Claude Code or Cowork.

The practical line is authentication. Signed in with your plan, claude -p uses plan limits. Run with a key from the linked Console organization, it draws on the credit.

The rules that will catch people

Three details deserve a note on the wall.

• They expire. The article says the credit "Doesn't roll over. Unused credits expire at the end of each billing cycle." Annual subscribers receive them monthly.
• One link, and you can't move it. "You can link only one Console organization, and each Console organization can receive credits from only one plan." Changing the linked organization requires contacting support.
• The balance is shared. Everyone holding an API key in the linked organization draws from it. The article suggests workspace spend limits in the Console for anyone who wants to fence off a project or a teammate.

The failure mode when the money runs out is benign. If the organization has purchased credits or auto-reload, usage continues on those. If not, "API requests stop until your next monthly credits arrive." Usage is never charged to the Claude plan. Monthly credits are spent before any purchased credits.

Anthropic also heads off one confusion: this is not the June Agent SDK credit, which the article says "isn't available."

Why it matters

For a Max subscriber who has been paying separately for API usage on small agents and scripts, this is up to $200 a month of that bill absorbed, with the condition that it is used inside the billing cycle. The design choice worth noticing is the boundary: the credit funds what you build on the platform, not more of Anthropic's own interactive tools. Anyone hoping to stretch Claude Code sessions with it will find that path closed.

Primary sources: Claude release notes, Monthly API credits for Max and Team plans, read 2026-10-20.