skip to content

say hello

Building something around AI memory, agents or on-device models? I read everything that lands in my inbox.

One account, one key, one balance for every top model.

private

A paid model router. Developers get one API key and one prepaid balance across Claude Code, Codex and a web chat, while a metering gateway prices every request against a live model catalog and settles it on a ledger.

highlights

  1. 01Go metering gateway with model aliases, request shaping, a usage tap and Redis hot state.
  2. 02Billing service on a Postgres ledger: accounts, magic-link sign-in, API keys, settlement, Stripe payments, plans, promotions, OIDC and scheduled jobs.
  3. 03A Rust CLI (aethene) that configures Claude Code and Codex in one step, with doctor, jobs and skills commands and npm, shell and PowerShell installers.
  4. 04Next.js site and dashboard (pricing, model pages, referrals, status, legal), all generated from a single catalog.yaml.
  5. 05Docker, Caddy, AWS CloudFormation and GitHub Actions deploys over OIDC. CI runs Go race tests against Postgres and Redis, cargo clippy/test and TypeScript checks.

built with

  • go
  • rust
  • next.js 16
  • postgres
  • redis
  • stripe
  • docker
  • aws

Routing and chat run on the open-source Bifrost gateway and LibreChat. Aethene adds the metering, billing, CLI, web and deployment layers.

↑↓ move↵ open⌘K toggle