Business & Funding

For $1.3 million a month, OpenClaw founder Peter Steinberger runs 100 AI agents that code, review PRs, and …

· May 16, 2026
For $1.3 million a month, OpenClaw founder Peter Steinberger runs 100 AI agents that code, review PRs, and …

What changed

Peter Steinberger and his three-person team run about 100 AI agents powered by OpenAI’s Codex for the open-source project OpenClaw. This setup drives their OpenAI API costs to roughly $1.3 million per month. These AI agents handle coding, reviewing pull requests, and bug finding at scale. Steinberger treats this astronomical spend as a research investment to explore software development unconstrained by token usage or costs.

Why builders should care

Spending over a million dollars a month on AI-driven development pushes the envelope on what automated programming can accomplish when cost is removed as a limiting factor. For builders, this reveals the kind of scale possible when deploying many AI agents simultaneously to manage complex workflows. It also exposes how AI can go beyond code generation to become an active participant in code review and bug detection, areas typically requiring deep human expertise.

The practical takeaway

For startups and engineering teams, OpenClaw’s model sets a performance benchmark that is currently cost-prohibitive for most. However, the experiment exposes what future software development pipelines may look like once large-scale AI usage becomes affordable. Operators should anticipate rising AI infrastructure costs as multiple AI agents become integrated into coding and quality assurance workflows. Those who figure out how to balance agent scale with cost efficiency will gain an operational edge.

What to watch next

Keep an eye on how OpenClaw’s approach evolves and if it inspires similar multi-agent builds in other development tools. Monitoring whether OpenAI or other providers respond with pricing or product changes to accommodate such high-volume users will also be key. Finally, watch for improvements in agent collaboration and autonomy that could make AI-driven coding systems both more efficient and deeper contributors to software quality.

AI Quick Briefs Editorial Desk

Stay ahead of AI Get the most important AI news delivered to your inbox — free.