Anthropic’s new AWS gateway puts SSO and spend caps in front of Claude Code

Claude apps gateway for AWS is a self-hosted access, spend, and policy layer for Claude Code and Claude Desktop. It ships inside the Claude Code CLI itself and runs as a single stateless container backed by PostgreSQL.
Auth goes through any OIDC-compatible provider. After a browser SSO handshake the gateway issues a short-lived token, one hour by default, so pulling someone from the directory kills their access within that window.
Admins set daily, weekly, and monthly spend limits per organization, group, or user. Blow through one and the gateway blocks requests until the period resets.
Configuration is a single YAML file, and the gateway routes requests to either Amazon Bedrock or Claude Platform on AWS. Model IDs match the Anthropic API: claude-sonnet-5, claude-opus-4-8.
Up next
Context compression cut 39% of tokens and saved almost nothing
Context compression cut input tokens 39% and saved essentially nothing: in a 25-task A/B on SR SWE bench, token mode busted the prompt cache 123 times, sending 6,062,098 tokens to full price. Lost cache discount: $16.37 against $15.41 saved.
Related stories
- Claude Enterprise adds admin analytics and cost controls
- Meta trims Claude Code use, yet spent over $105M in 28 days
- Barclays wants half its developers on Claude Code this year
- Samsung with Claude: “Design and verification shortened to 2 days from a month”
- UST puts Claude Code to work reading chip pinouts and writing regression tests
- Apiiro finds Claude Code secrets in 1,122 GitHub files
Comments
No comments yet. Be the first.
Join the conversation
Sign in to leave a comment. It appears after moderation.

