Engineering teams wanting background agents without running infrastructure; you pay tokens at cost plus 10%, with no seat fees.

Kenning's take

What it does

Ellipsis runs Claude Code or Codex agents in the cloud from YAML files kept in a GitHub repository. Agents can be triggered on a schedule or by GitHub, Slack, Linear or Sentry events to review code, fix bugs, build features or investigate production issues. Ellipsis handles the sandboxes, permissions, logging and budgets.

  • Agents defined in YAML in the repo
  • Runs Claude Code or Codex with any model
  • Cron, GitHub, Slack, Linear and Sentry triggers
  • Per-agent budgets and session logs
  • Managed cloud or deploy in your own AWS account

Who it suits. Engineering teams that want background coding agents without running the infrastructure.

Key facts

Made by
Ellipsis AI Inc.
Based in
United States (New York, per site footer)
How to get it
install on GitHub, self-serve; enterprise support packages optional
Works on
Web, GitHub, API, CLI
API price
tokens at cost plus 10%; CPU $0.142 per vCPU-hour and memory $0.024 per GB-hour on managed SaaS

Pricing

PlanPriceBilled
Managed SaaS$tokens at cost plus 10%usage
Your AWS account$10% of token cost plus AWS costsusage
Support Standard$5000month
Support Advanced$10000month
Support Premier$15000month

source

FAQ

How much does Ellipsis cost?

Agents cost the model tokens at cost plus a 10% fee, with compute billed by usage and no per-seat fees.

Which agents does Ellipsis run?

Claude Code or Codex, with the model chosen in each agent file.

Facts from the maker's own pages and the sources listed. Checked 2026-10-11.